Evolution - An Occasionally Coherent Ramble on Relief Pitching

Thursday, June 23 2022 @ 04:35 PM EDT

Contributed by: Magpie

Strategy evolves, surely. But I have a passing thought on evolution itself.

Evolution is often seen as a kind of progression, and indeed the process of natural selection rewards new kinds of efficiency, whatever they may be. But the very word "progression" already implies more than we should casually assume. Evolution generates change, but change is not necessarily improvement. We shouldn't simply assume that the new efficiency is better. Better for what?  Modern mammals, primates in particular, may be far more efficient and impressive than dinosaurs in many ways -  but 200 million years itself establishes a kind of marker that will be hard to match. I suppose it all depends on what you think the point of existence might be, and holy crap - has this veered off in an unexpected direction!

But things are not dissimilar with baseball. Strategy evolves, and those changes will tend to reward whatever new efficiencies can be found. They will be copied and adopted throughout the game. We all know what some of them are - power hitters, power pitchers, multiple short relievers, launch angles, defensive positioning etc. etc. That doesn't necessarily represent an improvement either. How has it made the game better, more fun to watch, more entertaining? It likewise depends on the answer to the question - w.hat is the point of baseball? For those involved in it, this is easy - the point is winning, and new efficiencies will almost always help achieve that narrow goal. But for the vast majority of us, the point might be something else. I'm not here for a long time. I am here for a good time.

This is going to be about relief pitching, which is something at the centre of baseball's strategic evolution over the past hundred years. We were discussing the subject here last week, stuff about how a modern manager should run his bullpen. This, after all, is our birthright as baseball fans. It's practically a sacred obligation. I'll bet they were complaining about the stuff Joe McCarthy was doing 80 years ago - in fact, I know they were - and he's merely the most successful manager who ever lived. And so this will be something of a continuation of that discussion. The Blue Jays at the moment appeared to be ricocheting like a pinball between either not having enough relievers ready to meet a crisis, or having to throw guys into games just to get them some work. There were dark suggestions that this might indicate some kind of failing on the part of the manager. I don't think so, I loudly cried - it's just a condition of the game. (I even dragged out the old Earl Weaver complaint, that "It's always the case that there's either not enough work available to keep everyone sharp or not enough arms available to bail you out of all this trouble. Always.")

Well, perhaps the manager could do a better job of employing his relievers with an eye on situational leverage in mind? it shouldn't be that hard. Just look at all those iPads in the dugout. Except that they already do. Most managers understand perfectly well the concept of leverage, and utilize their relievers accordingly.  Consider the relievers on your 2022 Blue Jays, and their Average Leverage Index: (All figures subsequently deployed are through the games of Tuesday 22 June. Even if I don't have this ready for the next off-day, I'll be damned if I update league and team numbers for the sake of a couple of extra games!)

Pitcher Lev. Index

Romano 2.651
Mayza 1.656
Garcia 1.644
Cimber 1.585
Phelps 1.052
Richards .983
Gage .979
Stripling .952
Merryweather .736
Thornton .529
Vasquez .227
Castillo .224

I really don't expect anyone would come up with something very different, if at all. (Well, would you? If you'd been making all the pitching changes these past ten weeks?) But nevertheless, I believe that the game is simply too fluid and unpredictable to establish much in the way of policies for doing things. Trust me, every manager who ever lived loves to have a set of rules he can follow, and does his best to discover any that can be used. The more of them, the merrier. It makes the job so much easier. I just play this guy when this is happening. I just do this thing when that is happening.  It reduces the number of actual decisions that must be made, in the moment, with the game swirling all around you. That's always a relief because there's always going to be more than enough decisions to make, pressing choices that will never go away. It even provides a way to deal with the endless second-guessing that goes with the territory. And so every manager pursues ("like a hamster on a wheel," I said) that elusive balance between having everyone sharp and everyone available. Every manager tries to find some order in an environment and in circumstances that change from day to day, from inning to inning, from pitch to pitch. It's an endless search for some kind of rational structure that can be imposed on that chaos. But the game will always be too damn fluid and unpredictable for anyone to manage the trick for very long.

Anyway. For more than a century, baseball managers have asked for less and less from the pitchers who start the games. This of course means requiring more and more from the pitchers who relieve those starters. This process has been unfolding gradually, but steadily and without respite, ever since they moved the mound back to 60 feet in 1893. I documented this, with pretty pictures and everything, back in 2007. And in the past fifteen years, all that's happened is the process has actually picked up a bit of speed. Furthermore, your modern baseball games seems to require an additional 20 pitches or so before it reaches a conclusion - which doesn't sound like much, but over the course of a season amounts to the workload carried by your fourth starter. It's a far bigger deal than we generally suspect.

SO - the entire job done by the pitchers is larger (all those extra pitches), and the share of the job being borne by the guys in the bullpen just keeps growing. And something else has changed over the last fifteen years.


Back in 2007, I started to suspect that John Gibbons seemed rather reluctant to employ a pitcher on three consecutive days. I checked the numbers, and indeed it was true. In 2007, Gibbons used a pitcher on three consecutive days just six times - only Ron Washington in Texas did it less often. Meanwhile, Joe Maddon in Tampa Bay was using one of his relievers three days in a row 23 times, a number of others four days in a row, and he sent Shawn Camp out there five straight days at one point. Well, they sure don't do that anymore. The entire American League, all 15 teams, in more than 1,000 games, have used a reliever three days running just 19 times. It's not even worth the effort of adding them all up.

Luckily, I was keeping track of how many times they used a pitcher on consecutive days - just two days in a row, that's all - so I have something I can share with you! (The second number, attached to team or player, indicates use three days in a row.)

KCR (55-3) - Staumont 8, Snider 8, Coleman 7, Barlow 7, Payamps 5-1, Garrett 5, Clarke 4-1, Cuas 4-1, Brentz 1, Speier 3, Griffin 1, Vizcaino 1, Abreu 1,
BOS (49-4) - Diekman 8-1, Scheiber 8-1, Strahm 8, Robles 5-1, Davis 4-1, Sawamura 4, Brasier 4, P.Valdez 2, Barnes 2, Danish 1, Houck 1,
TBR (49-3) - Adam 10-1, Thompson 7, Poche 6, Raley 5-1, Feyereisen 5, Kittredge 4-1, Wisler 3, Garza 3, Guerra 2, Armstrong 2, Beeks 1, Faucher 1,
CWS (41-1) - Hendricks 8-1, Foster 8, Ruiz 5, Graveman 5, Bummer 4, Sousa 3, Banks 3, Lopez 2, Burr 2, Crick 1,
TOR (38-1) - Romano 6-1, Garcia 5, Mayza 5, Richards 7, Cimber 5, Phelps 3, Merryweather 2, Borucki 2, Stripling 1, Thonton 1, Gage 1,

SEA (37-1) - Murfee 6, Sewald 5-1, Romo 5, Misiewicz 4, Swanson 4, Festa 3, Castillo 3, Munoz 3, Koch 1, Ramirez 1, Mills 1, Borucki 1,
NYY (32-0) - Holmes 7, Chapman 5, Castro 4, Peralta 4, Green 3, Luetge 3, Loaisga 3, King 2, Marinaccio 1,
BAL (31-0) - Lopez 8, Bautista 7, Perez 5, Tate 4, Krehbiel 3, Fry 1, Baker 1, Gillaspie 1, Vespi 1
OAK (31-0) - Jackson 5, Puk 4, Acevedo 4, Trivino 4, Snead 3, Kolarek 3, Moll 3, Jimenez 2, Markel 1, Pruitt 1, Grimm 1,
DET (29-0) - Soto 6, Lange 5, Fullmer 4, Chafin 4, Vest 4, Jimenez 2, Foley 2, Peralta 1, Garcia 1,  

CLE (25-4) - Clase 7-2, Shaw 6-2, Stephan 5, Hentges 3, Sandin 1, Morgan 1, DeLos Santos 1, Gose 1,
TEX (24-0) - Barlow 7, Martin 4, Bush 3, Santana 2, Holland 2, King 2, Patton 1, Sborz 1, Moore 1, Tinoco 1,
HOU (22-0)- Neris 7, Montero 4, Pressly 3, Maton 3, Taylor 2, Mushinski 1, Blanco 1, Stanek 1,
LAA (23-0) - Loup 7, Iglesias 6, Tepera 3, Barraclough 2, Quijada 2, Mayers 1, Bradley 1, Herget 1,  
MIN (13-2) - Smith 5-1, Thielbar 2, Pagan 4-1, Moran 1, Duran 1,  

This seems a pretty clear attempt to spread the load around as many arms as possible. Kansas City has the worst pitching in the league - it would seem to follow that Mike Matheny would be going through one reliever after another. And so he should, I suppose. I certainly believe that any manager, if he's at all good at his job, will respond to the skills of personnel he has on hand rather than impose his own preconceptions of how the game should be managed. But hot on Matheny's heels are Alex Cora in Boston and Kevin Cash in Tampa Bay, running two of the better starting rotations in the AL No one goes to the bullpen more often than Cora, but Cash is hot on his heels, along with our own dear Charlie Montoyo. It's not really working for Cora - AL relief pitchers give up fewer runs (3.68 ERA) than starting pitchers (3.97 ERA) but the opposite is happening in Boston, whose starters have been better than his relievers (3.46 ERA for the relievers, 3.85 for the starters.)

Cash's staff has a more normal difference in performance from the starters and relievers. No bullpen crew has worked more innings than Tampa Bay's, no group of starters has worked fewer innings. Kevin Cash is as relentlessly logical as any manager in the game - his starters are good, but his relief pitchers are better. SO this is what you get. Cash has a method, and when he hooks his Cy Young winner working on a shutout in the sixth inning of a World Series deciding game, only to see his bullpen lose the game- well, he can at least tell you why.  Dusty Baker seldom seems so committed to logic, but he could surely sympathize. He once pulled a starter working on a shutout in the seventh inning of a game that would have won the World Series, only to see his bullpen lose the game. There's a lot of ways to win, and at least as many ways to lose.

The two managers who make the fewest pitching changes are two of the three AL managers who were managing way back in the 1990s - Baker in Houston and Terry Francona in Cleveland.  It's definitely not because they have sub-par bullpens - Houston's relievers have posted a 2.59 ERA, best in the league, and Cleveland comes next at 2.86 (the worst, naturally, works for poor Mike Matheny. Dude can't win no matter what he does.) Baker, however, also has one of the league's best group of starting pitchers (second only to New York) so his approach makes some sense. Cleveland's relievers, however, have been quite a bit better than the starters, in excess of the normal spread.

But this is a Digression! While the whole subject is of considerable interest (to me, anyway!) it's not something I had even intended to examine. It just rose up in the course of my investigations, which began with something quite different.


You see, what I actually thought of, in the moment while we were discussing the subject, was that we hadn't even considered the fact that modern managers seem to prefer giving their relief pitchers a clean inning. That would surely be another factor to additionally complicate bullpen deployment...

And that's what stopped me. They certainly seem to prefer giving the reliever a clean inning - but do they, really? How would I find out? I didn't know.  I hadn't even bothered to consider whether it would be significant or not. Clean Innings! I simply had to know. So down the rabbit hole I went, looking for clues. I knew, alas, that I would not find the footprints of a gigantic hound. But maybe there would be something!

I didn't know how to find the answers I sought and I immediately made a serious wrong turn into Inherited Runners, thinking that this might suggest whether or not managers were trying to give their relievers clean innings. I even got the good folks at bb-ref.com to confirm for me that - just as I suspected, but wasn't certain - the total number of Inherited Runners for any team (103 at the moment for the 2022 Blue Jays) may include some double counting. The actual number of runners inherited by the bullpen is likely to be a little less than the total inherited by the individual relievers. (This is because one reliever may inherit a baserunner and pass that same runner along to a second reliever.) It's a small thing. It is the case that the number of Inherited Runners faced by a team's bullpen has declined over the past 45 years or so. It's not as dramatic as some of the other changes in bullpen use, but it's a real thing and the only explanation I can think of would be that the quicker and quicker hooks have made a little bit of difference. But it's not a subject that really tells us much about the issue that concerned me (clean innings!) although it's not without its own interest.

And then - oh happiness! - I finally noticed that the splits for relievers actually included the number of times there were Runners on Base and the number of times the Bases were Empty when a reliever entered the game. Oh frabjous day! Callooh callay! I chortled in my joy.  Now we're getting somewhere, I said. ("We are?" came the reply from across the room.) And so the copying and pasting began, for every AL season going back to 1977.  I was going to go all the way back to the first DH year - but I decided to do the same thing for the Blue Jays, in the event that the parallel stories proved of interest. And they did. And because they did, I also conducted some forensic examinations of how relievers were used in a number of Blue Jays seasons. As I've made my clear on many occasions, I have no life. Even so, I planned to just look in on the Jays at five-year intervals and chose  the years ending in '2 and '7 (the first and the last!), but the years around 1990 required additional examination.

How shall I tell this story, share the tiny nuggets I have gleaned?  With a Data Table, obviously. But in my old age, I seem to have developed some strange feeling of... compassion, possibly? - for my readers. Naturally, I hope it goes away as soon as possible. But I think I shall break the Data into smaller, more digestible chunks. And not include all the raw numbers - just relevant percentages, and stuff like that.

Do modern managers try to give their relievers a clean inning? You bet they do. Do they ever. Even though they're using more and more of them, even more they're asking them to pitch more and more of the team's innings - they very much prefer to bring in a new pitcher to begin the inning fresh. They would much rather make the pitching change at a moment of their own choosing. As opposed to the old way, when managers brought in a new pitcher because they simply couldn't leave the guy on the mound out there a minute longer. When it wasn't nearly as much of a  choice at all.

Yet another digression! One of the things I wanted to know was whether managers made much of a habit of changing pitchers in the middle of an inning with no one on base. We saw it just the other day, when Tony LaRussa brought in Tanner Banks to face Raimel Tapia with the bases empty and two outs (in a 9-1 game, yet.) I think we're all agreed that the practise is an abomination - I hereby dub it the Pointless Pitching Change - and I suspect it's provided some of the motivation for the three batter rule. But to be fair, sometimes there is a reasonable point to the Pointless Pitching Change. If you've got a one-run lead, a RH on the mound, and Bryce Harper coming to the plate - well, I think  I'd be looking for the LH too, and I wouldn't care how many outs I had in the inning. This is something I only actually dug into for those selected Blue Jays seasons that I investigated in excessive detail, because you need to go to the Game Logs for individual pitchers. I may not have much of a life, but please! And it might not that big a deal. Cito Gaston never made more than 6 such pitching changes in any of his first five years on the job. Still, by 1997 even he was making 27 Pointless Pitching Changes in a single year. John Gibbons and John Farrell would both crack 30 of them in a season, and the highest figure I've come across so far was 39, from the Buck Martinez-Carlos Tosca tandem in 2002. End of digression, for now.

How shall I organize this great whopping mess of material? When in doubt, chronological order can be your friend.
            Games    RP per  RP % of  Bases    IR                           RP per  RP % of   Bases
Year    Lg   Avg.    Game    Innings  Empty %  Game        Team   Games    Game   Innings   Empty %
1977    AL    120    1.30    28.2%    33.1%    1.98        Toronto    121    1.16    26.8%    27.3%
1978    AL    116    1.21    27.1%    31.0%    2.01        Toronto    126    1.32    29.0%    25.8%
1979    AL    122    1.34    28.5%    31.8%    2.04        Toronto    118    1.20    28.1%    30.3%

These were the Olden Days, and I'll also take a moment to comment on the figures I'm providing. The Games, as you have surely guessed, are the number of games that required the use of a relief pitcher. (This figure will be pro-rated to a 162 game schedule for the several shortened seasons since then.) Complete Games had tumbled from about 80% in 1905 to about 50% as the 1920s arrived. They continued to decline, gradually but inexorably until by 1977 starters were completing roughly one game out of every four. Relief Pitchers per Game are the number of relievers required per every game on the schedule. The number of relief pitchers used in a game had first crossed 1 per game for individual teams as early as the 1920s, but the AL as a whole didn't cross that threshold until 1948 (it had been surpassed as early as 1934 in the NL.)

By 1977, AL teams were using an average of 1.3 relievers per game, and those relief pitchers were working 28.2% of the team's innings. And they usually came into the game when trouble was brewing - if the bases were empty just 33.1% of the time, that obviously means that there were runners on base 66.9% of the time. (I've provided the number of Inherited Runners the bullpen faced per game mainly because it's where I started.)

Roy Hartsfield was an old-fashioned type of manager even in his own time. Despite his dreadful starting rotation, he still used fewer relievers and for fewer innings than the rest of the AL. Even more than the rest of the league, he had to be forced into changing pitchers. There were runners on base 75% of the time before he would go to the pen. Hartsfield had just three relief pitchers getting steady work - Johnson, Vuckovich, Willis - and they appeared in 43, 45, and 46 games. Oddly enough, Hartsfield was fonder of the Pointless Pitching Change than I might have expected - 13 pointless changes seems quite a bit under the circumstances. On eight of those occasions, Mike Willis came in, so one assumes there was a scary LH batter coming up.
       Games    RP per  RP % of  Bases    IR                 RP per  RP % of   Bases
Year    Lg   Avg.    Game    Innings  Empty %  Game        Team   Games    Game   Innings   Empty %

1980    AL    122    1.39    29.0%    35.6%    2.01        Toronto    123    1.77    33.3%    31.5%
1981    AL    126    1.45    29.2%    36.0%    2.04        Toronto    130    1.78    31.5%    39.7%
1982    AL    130    1.46    30.3%    38.1%    1.87        Toronto    121    1.36    28.2%    45.5%
1983    AL    129    1.44    28.9%    39.0%    1.85        Toronto    119    1.59    28.1%    38.9%
1984    AL    134    1.54    29.1%    44.1%    1.76        Toronto    129    1.58    25.3%    40.1%
1985    AL    136    1.65    30.8%    43.0%    1.88        Toronto    143    1.96    29.9%    48.7%
1986    AL    137    1.66    30.5%    43.1%    1.89        Toronto    147    1.78    32.1%    48.6%
1987    AL    135    1.73    31.2%    41.7%    2.03        Toronto    144    2.07    33.3%    37.8%
1988    AL    136    1.65    29.0%    44.1%    1.87        Toronto    146    1.81    33.2%    40.5%
1989    AL    143    1.81    31.3%    46.9%    1.81        Toronto    150    1.71    32.9%    50.9%

The decline in number of Complete Games is picking up speed - AL teams were averaging 40 CGs per season at the beginning of the decade - they were averaging just half as many by the end. The number of relievers per games rises as the CGs fall, as does the number of innings absorbed by the bullpen. And more and more often, those relievers are entering the game when there's no one on base. By the end of the decade, almost half the relief appearances came with no one aboard.

And that brings us to the Blue Jays of the 1980s. Bobby Mattick ran the ship for the first two years, and immediately made much more frequent use of his bullpen than Hartsfield ever had. But Bobby Cox took over in 1982, and Bobby Cox's great gift as a manager, in both Toronto and Atlanta, was his way with starting pitchers. The Blue Jays began their climb to goodness on the backs and arms of their young starters - Stieb, Clancy, and Leal at first. Cox leaned on them very heavily - he was one of the last managers to employ a four-man rotation, and the number of relievers the Jays used fell sharply from what it had been under Mattick. Dave Stieb led the league with 19 CGs.  Cox tended to bring them into a game with the bases empty more often than the other AL managers. Dale Murray was his first relief ace, and there were more often men on board when Murray was needed (34 of his 56 games) - but Roy Lee Jackson got almost as many bases Empty situations as games with baserunners, and Cox seems to have actually made an effort to give his third major reliever (Joey McLaughlin, whom he knew from Atlanta) a clean inning - in just 17 of his 46 games were there men on base. Cox and the bullpen were, as always, an erratic thing. He switched to a five man rotation in 1983 and started using as many relievers as the other AL managers, even if they didn't absorb quite as many innings.  in 1984, the number of relievers used and their innings dropped rather sharply. Then in 1985, his last year here, he used more relievers for more innings than ever before in his Toronto career and almost made a point of giving them a clean inning to work with. His 48.7% of Bases Empty relief entries was well above the league average, well above anything seen in Toronto.

Cox moved on to Atlanta, of course, and Jimy Williams took over in 1986. Immediately, the Blue Jays begin using more relievers and for more innings than the AL average. While Williams wasn't using as many relievers as Cox had in his final season, he did use them for significantly more innings. The reason is obvious - Mark Eichhorn pitched 157 IP of kickass relief in 69 games (which tied Gary Lavelle's franchise record.) These weren't clean innings - there were runners on base in two-thirds of them. But Lamp and Caudill, having poor seasons, were more often given a clean inning than trusted to deal with runners in place - between those two and Henke (whose appearances were split evenly between Runners aboard and Bases Empty), the Jays were giving their relievers a greater share of Clean Innings than the rest of the league, although it was clearly a growing trend all over.  But Williams' was a compulsive juggler and meddler, and he was doing everything differently a year later - Eichhorn made 20 more appearances and pitched 30 fewer innings. Jeff Musselman gave him a viable LH option, and like Eichhorn was used much more frequently when there were runners on base. The number of relievers used rose well above the league average, while the number of clean outings plummeted below. 1987 was a year for offense, with a record number of HRs hit. The long balls fell off sharply the next year, runs scored fell along with them, and no one in Toronto or the rest of the AL saw the need to use as many relievers as they had in 1987. And then, in 1989, the Blue Jays crossed a threshold, with more than 50% of their relief appearances coming with the bases empty. What brought this on?

Well, there was a new guy in charge. And if you thought 1989 was different - just you wait.

            Games    RP per  RP % of  Bases    IR                           RP per  RP % of   Bases
Year    Lg   Avg.    Game    Innings  Empty %  Game        Team   Games    Game   Innings   Empty %

1990    AL    146    1.96    31.9%    50.1%    1.84        Toronto    156    1.96    33.4%    66.9%
1991    AL    147    2.10    32.6%    51.3%    1.85        Toronto    152    2.14    30.6%    64.0%
1992    AL    145    2.06    30.8%    53.3%    1.82        Toronto    144    1.75    28.0%    75.7%
1993    AL    147    2.19    31.4%    51.8%    1.95        Toronto    151    2.12    30.7%    64.2%
1994    AL    147    2.25    31.6%    54.7%    1.88        Toronto    145    1.92    30.3%    71.5%
1995    AL    150    2.32    33.7%    54.8%    1.92        Toronto    144    1.84    29.5%    65.7%
1996    AL    150    2.36    33.8%    54.0%    1.94        Toronto    143    1.87    30.3%    72.3%
1997    AL    153    2.45    33.3%    58.4%    1.82        Toronto    143    2.07    26.9%    64.9%
1998    AL    152    2.49    32.4%    57.3%    1.87        Toronto    153    2.36    27.4%    56.5%
1999    AL    154    2.52    34.7%    58.4%    1.87        Toronto    148    2.33    32.9%    59.9%

In 1990, the AL as a whole crossed (barely) the 50% threshold for bringing in relievers with the Bases Empty, and Cito Gaston's Blue Jays led the way and then some - Gaston's relievers entered the game with the bases empty two-thirds of the time. This was the highest such figure by any AL manager since Ralph Houk of the Yankees way back in 1971. The league as a whole hadn't been over 50% since 1972. And I think you can all see right away what had happened back then - in 1973, the AL adopted the Designated Hitter. As a result, AL managers were no longer bringing relievers into the game because they'd just pinch hit for the pitcher, which would obviously give that reliever a clean inning to start with, AL teams had averaged almost 1.5 relievers per game (1.47) and 55.3% of the time they came on with the bases empty. This is quite similar to what was going on in the NL (1.44 and 57.2, to be precise.) But in 1973, with the need to pinch hit for the pitcher gone, the use of relievers in the AL dropped to barely more than 1 per game (1.15) and almost 70% of the time they were coming into a game with runners on base (they came in with the bases empty just 30.5% of the time.) The National League continued to move in the opposite direction - more and more relief pitchers, who more and more often were coming in with the bases empty, more often than even the requirements of pinch hitting mandated. The introduction of the DH set this process back for a generation in the AL, though I hasten to add that I do not regard that as a bad thing.

Gaston certainly wasn't the first AL manager to give more than half his relievers a clean inning in the DH era. Gene Mauch and Sparky Anderson, among others, had been known to go there as well. But only Dick Howser in Kansas City seemed to be as utterly committed to the strategy as Gaston. Howser's teams in New York and Kansas City would lead the AL in Bases Empty pitching changes in each of his five full seasons as a manager. His 1984 team posted what would stand as the highest percentage of any AL team in the DH era until Gaston came along. Likewise, Gaston's relief crews led the league in percentage of Bases Empty appearances, often by a hefty margin, in each of his first nine seasons on the job (Phil Garner's Brewers just edged past him in 1997.) Gaston was still managing this way when he returned to the dugout a decade later, still giving his relievers a clean inning about 70% of the time. By this point, the rest of the game was beginning to catch up to the Great Innovator - Gaston's 2009 and 2010 teams were merely close (second and third) to the top of the heap by now.

Gaston was especially dedicated to the idea of giving his ace reliever a clean inning to work with. He basically had a four man bullpen in 1990 - all four men came in more often with the bases empty but Henke and Ward in particular were much less likely to inherit anyone else's runners.  He did things the same way in 1991, except now he had a fifth reliever, LH Bob MacDonald. Gaston handled MacDonald much differently that first year. The southpaw was much more likely to be called on with there were runners on base, doubtless seeking a platoon advantage. Things began to change in 1993. The pitching on the second WS winner was nowhere near as good as that on the 1992 team and Gaston was forced to make almost twice as many pitching changes to deal with enemy baserunners. Eichhorn and Castillo bore the brunt of this dirty work - but even that pair still came into a game more often with the bases empty. And as always Gaston made a point of giving his ace reliever (Ward alone, by this time) a clean inning to work with. Gaston had some pretty ordinary staffs after that, but he still continued attempting to give his relievers a clean inning as often as possible, certainly more often than the other AL managers, and as always made a special effort to do so with his bullpen ace, whoever that might be (it was changing every year by now, from Hall to Castillo to Timlin to Escobar.)

In the game as a whole - Complete Games continued to fall, the number of relievers per game continued to grow, the number of innings they were soaking up continued to grow.  I expected the 1990 spring training lockout to have made a larger impact on the game as a whole - like the one this spring, it wiped out much of spring training (an agreement wasn't reached until 19 March) and it was obviously impossible for teams to get their starters ready in time for Opening Day, even if it was moved back a week. Gaston's Blue Jays would set a major league record (to that point) for fewest Complete Games (6) in a season. (It's been broken since!) But the lockout doesn't seem to have nearly as much effect league-wide or game-wide. The same processes continued, at about the same rate. The decade is notable mostly for the offensive explosion that began in 1994, which certainly wasn't something that discouraged managers from making pitching changes..

Time for a brief digression on the Pointless Pitching Change. Bobby Cox seemed to grow more willing to indulge in them as the team got better - he made just 10 of them in his first season, but was up to 24 of them while winning 99 games in 1985. Jimy Williams peaked at 22 in his best season, when the team won 96 games in 1987- he was all the way down to 7 the next year. Cito Gaston would come close to dispensing with them altogether, and he did so while his team was winning. Gaston would make just 5, 6, 8, 1 (yes!), and 4 of them in his great initial run from 1989 through 1993, while his club played .573 ball.

       Games    RP per  RP % of  Bases    IR                 RP per  RP % of   Bases
Year    Lg   Avg.    Game    Innings  Empty %  Game        Team   Games    Game   Innings   Empty %
2000    AL    154    2.52    34.8%    59.7%    1.80        Toronto    147    2.40    33.4%    60.8%
2001    AL    154    2.51    33.7%    62.8%    1.61        Toronto    155    2.91    34.2%    62.4%
2002    AL    154    2.45    32.6%    64.6%    1.53        Toronto    156    2.85    34.8%    64.6%
2003    AL    154    2.51    33.7%    64.7%    1.57        Toronto    148    2.74    31.9%    56.5%
2004    AL    156    2.62    34.1%    64.1%    1.66        Toronto    155    2.68    34.4%    68.0%
2005    AL    156    2.61    32.8%    65.0%    1.57        Toronto    153    2.67    34.5%    60.9%
2006    AL    157    2.71    34.3%    64.8%    1.62        Toronto    156    2.98    37.0%    64.1%
2007    AL    157    2.78    34.2%    64.6%    1.70        Toronto    151    2.59    31.1%    60.0%
2008    AL    157    2.77    34.3%    67.5%    1.60        Toronto    147    2.60    29.4%    68.9%
2009    AL    157    2.80    34.6%    66.7%    1.58        Toronto    152    2.75    33.6%    73.0%

The first decade of the millennium was pretty chaotic for the Blue Jays - three mid-season managerial changes? five different managers? - but let's look first at some trends in the game as a whole. In 2002, Jerry Narron's Texas Rangers allowed 882 runs while going 72-90, and in a desperate attempt to stop the bleeding Narron called on 487 relief pitchers. Not only was this the most relievers ever used in a season, it made his Rangers the first team ever to average three relief pitchers per game. A forgotten pioneer! It would take more than a decade for the league to catch up, but catch up it would. Until 2007, no team had ever gone through a season without getting a single complete game from one of its starting pitchers. Not a single one.  Ever. But in 2007, no less than three teams -  Florida, Washington, Texas - crossed that threshold. The future was picking up speed. A bare handful of games were now being completed by starters. Maybe one per month.. The surge in offense from the previous decade began to abate, but managers continued to increase the number of relievers being used and the number of innings they were being asked to pitch. But the number of innings was not growing as rapidly as the number of relievers - just like starting pitchers, individual relievers were being asked to pitch fewer innings than they had in past seasons.

The Ricciardi era in Toronto was of course a saga of frustration, chaos, and disappointment. I think the experience was best described in the last words spoken by Johnny Rotten at the Sex Pistols final concert at Winterland in 1978. By far the best thing about the Jay, apart from Roy Halladay, was the entrance into our lives of this mighty place right here. The first posts at Batters Box appeared in October 2002. I myself stumbled across this joint midway through the dreadful 2004 season. It was fun, it was a distraction from the pain and dread on the field. I wanted to get in on the discussion, and everyone seemed to be using an Internet nom de plume... so I've been Magpie from that day forward. It suits me, I think - like my master Autolycus, the magpie is a snapper-up of unconsidered trifles. He is easily distracted by whatever bright and shiny object catches his eye, and he dearly, dearly loves to cause a little Trouble without Consequences. I was invited to come on board that off-season, and they've been stuck with me ever since. You're welcome.

Enough of that nostalgia, and on to this nostalgia. Because there were some things happening on the field that come under our purview here. In 2001, rookie manager Buck Martinez summoned no less than 471 relief pitchers out of his bullpen. This utterly obliterated the old franchise record of 388, set under Jim Fregosi's management the previous season. John Gibbons' pen would surpass that mark in 2006. With different men in charge, the old emphasis on giving relievers fell and was usually a little below league levels until Gaston returned in 2008. But we still see the usual emphasis on giving the relief ace a clean inning, and letting the other relievers deal with brewing trouble.

            Games    RP per  RP % of  Bases    IR                           RP per  RP % of   Bases
Year    Lg   Avg.    Game    Innings  Empty %  Game        Team   Games    Game   Innings   Empty %

2010    AL    155    2.71    32.1%    68.1%    1.48        Toronto    157    2.81    33.5%    69.9%
2011    AL    155    2.69    32.2%    68.7%    1.43        Toronto    155    2.93    33.9%    69.0%
2012    AL    157    2.91    34.2%    68.4%    1.52        Toronto    157    3.06    36.6%    70.7%
2013    AL    158    2.89    34.3%    67.7%    1.57        Toronto    158    3.01    38.1%    69.0%
2014    AL    158    2.98    33.9%    68.6%    1.55        Toronto    159    2.77    33.6%    63.9%
2015    AL    157    3.02    34.5%    68.4%    1.57        Toronto    155    2.90    33.0%    71.2%
2016    AL    159    3.01    36.1%    71.1%    1.43        Toronto    162    3.01    31.8%    72.9%
2017    AL    160    3.19    38.2%    69.2%    1.60        Toronto    160    3.57    40.7%    68.2%
2018    AL    160    3.31    40.9%    72.1%    1.54        Toronto    162    3.64    41.5%    72.2%
2019    AL    160    3.40    44.1%    73.3%    1.52        Toronto    161    3.65    50.6%    72.3%

This would be the modern world, and the same process we've watched over the preceding 40 years continue to grind away. The number of relief pitchers used per game continues to grow, moving above 3 per game for the league as a whole for the first time in 2015. And the number of innings absorbed by those relievers grows along with it, and by 2018 relief pitchers are working 40% of a team's innings for the first time in AL history. And more than ever before, AL managers are making a real effort to give this host of relief pitchers a clean inning on which to start their work. These trends unfolded on a parallel track in Toronto, although both John Farrell and John Gibbons Mark II tended to use more relievers for more innings than the average AL team. Farrell's 2012 team was the first Blue Jay squad to average 3 relievers per game. Gibbons cut back on his use of his relievers in 2014, year two of his second term and for three seasons was asking less of his bullpen than the rest of the AL. And then, in 2017, everything gotr very weird. Gibbons called on an enormous number of relievers - 578 of them, obliterating the old team mark by almost 100, just as Buck Martinez in 2001 had obliterated the old team mark. They worked more than 40% of the team's innings, also a first in franchise history. This was largely brought on by the starting pitcher catastrophe - Happ and Sanchez got hurt, Liriano was traded away with the team out of contention - and the ensuing carnage demanded a great deal from the bullpen. Gibbons still did his best to give his relievers clean innings to work with, and basically insisted on it for his closer, Roberto Osuna.  Dominic Leone more often had a mess to clean up than a fresh inning, but he was the only one. And 2018 was more of the same - the rotation was an even bigger caytastrophe, this time because everybody simply sucked. Except Happ, who was traded at the deadline. Gibbons needed even more relief pitchers to make it through the season, 590 of them working 41.5% of the team's innings.

That was it for Gibby, and the Montoyo era began in 2019. And while in his first season he used almost exactly the same number of relievers as Gibbons had - 591, just one more, but enough to set yet another new team high - Montoyo's 2019 bullpen actually worked more than half of the team's total innings. This, surely, was terra inconnu? I suppose, but what had happened is that Montoyo had come from Tampa Bay and brought along with him some of those filthy Tampa Bay ways. In this case, the use of The Opener. By my count, 23 games were started by a relief pitcher, who would work no more than two innings and hand the rest of the game over to the rest of the bullpen. And just like everyone else in baseball, he was going out of his way to give those relievers a clean inning to work with. By now, barely one in four relief pitchers who come into a game are expected to have to sully their hands with the dirty task of stranding someone else's baserunners.

            Games    RP per  RP % of  Bases    IR                           RP per  RP % of   Bases
Year    Lg   Avg.    Game    Innings  Empty %  Game        Team   Games    Game   Innings   Empty %

2020    AL    160    3.45    45.5%    70.6%    1.57        Toronto    162    3.77    51.3%    66.4%
2021    AL    161    3.33    42.9%    70.6%    1.45        Toronto    161    3.31    43.4%    68.9%
2022    AL    160    3.43    42.7%    70.4%    1.49        Toronto    162    3.70    41.7%    71.7%

These are the plague years, and we simply haven't had a truly normal season yet. Maybe next year. The 2020 season was chopped to a mere 60 games. Last year, 2021 was in many ways a process of recovery from the previous year's short season. We were warned in advance that there might be lots of problems with pitchers ramping back up to a regular workload, and sure enough - there were. And this year has repeated the story of 1990, with spring training largely wiped out by a lockout, and teams still in the process of ramping up into regular season .form while the games actually counted. With respect to this year's pen, we can see that while Cimber and Phelps are generally Montoyo's choices when there are runners aboard, he still manages to give them a clean inning more often than not. Garcia and Romano hardly ever have to deal with someone else's runners. He's disturbingly fond of the Pointless Pitching Change and may very well equal the 34 of them John Gibbons inflicted on us back in 2017. I think Carlos Tosca's mark of 39 is not yet in any danger (if that's even the record - I haven't checked every Jays season for it.)