A Solution for September Roster Insanity
So, things might be getting out of hand…
Also, Bud Norris has been activated. #Angels have 14 relievers.
— Jeff Fletcher (@JeffFletcherOCR) September 6, 2017
A lot of folks inside and outside the game don’t care for September roster expansion. After all, why play the game one way for five months, with one set of roster rules, and then in the most crucial month of the season change the limits of rosters?
If we were drawing up a rulebook from scratch, if we were creating an entirely new collective bargaining agreement, I don’t think we would arrive at this idea of September roster expansion. But here we are. Again
Those in favor of the September rules may note that many players make their MLB debuts in the final month of the regular season after the minor-league seasons have been completed and that the MLBPA has legitimate concerns about service time. The MLBPA should have concerns about service time given the way teams have manipulated the debuts of young players to enjoy another year of club control and/or to avoid the additional year of arbitration eligibility which Super 2 status entails.
So there are legitimate arguments on both sides. There’s room for compromise, though, and I think a fairly reasonable exists: the addition of another year-round roster spot (increasing the active roster to 26 players) and a corresponding reduction of the September roster limits. This would address the problem in a number of ways.
An additional spot on the active roster would almost always be utilized for another pitcher and that could represent a significant benefit for teams — not only by facilitating greater match-up flexibility but also by helping to keep arms fresher and more healthy. Starting pitchers’ workloads continue to decrease. In 2015, there were 27 pitchers who reached the 200-inning threshold, the lowest of the expansion era. In 2016, that figure dipped to 15. Nonetheless, pitcher injuries remain at or near record levels in recent seasons.
While the 10-day DL has been used by some teams as a rotating roster spot — as has the final bullpen spot on the roster — a 26th arm would be a more even-handed way to create pitching depth. (Or a team could still use the DL, like, say, the Dodgers, and create a 27th man.)
By adding a 26th man, the union is guaranteed a service time bump from April through August for 30 teams, roughly 4,500 days of total additional service time prior to September. There are no such guarantees with the current system.
By adding a 26th man, perhaps the embattled middle class of player — one who’s had trouble finding work in recent springs — would enjoy work more quickly. That’s 30 new openings (job creation!) at a time when the owners’ share of revenue continues to increase.
By adding a 26th man, by reducing the September roster spots dramatically from a maximum of 40, it ensures that the game is played in its most important month in roughly the same way it’s played over the first five.
While it’s hard to get two parties to agree on anything, this seems like something that could have owner and player bipartisan support, and the idea — or something similar — was close to becoming a reality during last winter’s CBA negotiation, according to the Associated Press.
Interestingly, it was the union that backed away.
“We thought we were going to make an agreement, had a tentative agreement,” baseball commissioner Rob Manfred told The Associated Press on Friday, “but nothing’s done until it’s done. … There were very mixed opinions on the club side, as well. Maybe we were going too far, too fast.”
Said MLPBA chief Tony Clark last winter:
“I don’t know that there was an agreement to do it and that it came apart. There was a lot of dialogue over the course of this round where we were moving in a direction that we inevitably weren’t able to agree to,” he said. “It simply got to the point with all of the moving pieces that were part of the conversation during the course of the year that we lived to talk about it another day.”
The MLBPA hasn’t been winning many meaningful battles lately. The average age of major-league players has decreased since PED-testing began, which means hundreds of seasons once played by veteran players have been replaced by the production of younger, cheaper players. Owners’ profits are up. Again, the 26th man is a job-creation opportunity that would also help the integrity of the game in September, when the product should be at its best for the consumer.
Baseball could use a 26th man in concert with smaller September rosters. Instead, the Angels have 14 relievers.
A Cleveland native, FanGraphs writer Travis Sawchik is the author of the New York Times bestselling book, Big Data Baseball. He also contributes to The Athletic Cleveland, and has written for the Pittsburgh Tribune-Review, among other outlets. Follow him on Twitter @Travis_Sawchik.
Well, the Angels actually need 14 relievers at the moment, as they don’t seem to have any real starters currently.
That might be one thing, but Scioscia being lunatic is another. I had the great pleasure of watching the A’s play the Angels recently and saw ol’ Mike use 12 pitchers in an 11 inning game that took 4 hours and 38 minutes. Of those 12 pitchers 5 of them faced 2 batters or fewer, 3 of them didn’t even record an out!
Hey, someone had to replace Tony La Russa
I think MLB should take from the NHL and have “healthy scratches.” Let there be 40 men on the major league roster, but only 25 on the gameday roster posted with the lineup card.
I think this is really the way to go. This way you don’t mess with basic formula of 25 players in a game, you don’t solve the problem with some weird special case thing where you get an extra pitcher on a weekend of a double header, and you don’t have some big change to the roster come September.
It checks all the important boxes as far as I’m concerned.
Remember that there are 4 starters who are not relevant on game day. This essentially expands rosters by 4.
Yeah, so maybe 20 active players on game day or something like that.
So you allow that any player who started a game may not be a healthy scratch for the next,4? 5? days. An injured pitcher may, of course, be placed on the actual DL.
…until some team gets the bright idea to have a dedicated “opener” pitcher, who pitches the 1st inning in a majority of games, and the would-be starter comes in to start the 2nd instead. They’ll be forbidden from scratching the opener, not that they were planning on it, but that first guy out of the bullpen becomes perfectly scratchable.
Keeping the existing April-August roster rules for those months is fine, and “40 with 15 pregame scratches” during September call-ups is a reasonable modification of that. No exceptions about not being able to scratch a starter or anything, such that you’re incentivizing teams to change their pitcher usage patterns in order to “flout” the rule. The four inactive starters would be a routine sight under that system, effectively giving teams a 29-man roster in September, and that’s fine. 29 is a modest increase over 25, but it’s still a lot lower than 40.
I think it would be simpler to set an “active” roster of, say, 22 players per game. Assuming teams designate their last 4 starters as not active, that increases the effective roster by 1 player. During the regular season maybe you allow teams to have their reserve roster up to 26 players, and in September that reserve can increase up to the full 40.
any pitcher to throw 60 pitches or 4 innings in the past 5 days, or a pitcher who has averaged 60 pitches in their last 4 “games” and who pitched 40+
The NFL has 53 man rosters but only 46 eligible on game days.
I like the “healthy scratch” idea too. We can just have the 40-man roster the total roster, with the people on the 40-man either game-day active, healthy scratches, or assigned to a minor league affiliate. The 40-man roster then gets to do double-duty.
The one thing I’m curious about is how this would interact with the 10-day disabled list. I just would want to make sure there weren’t any crazy unintended consequences there.
A 25-man roster doesn’t stop a team from having 1-2 starting pitchers and 11-12 relievers, which still allows a 12-13 man pitching staff. If your idea was implemented, there should also be a rule that teams are only allowed 7-8 relievers per game. Teams would occasionally be penalized if their last reliever was lost due to injury, but we already see position players pitching when teams try to save their bullpens anyway.
This causes an issue in extra inning games or if a SP gets hurt. I’d rather have the current system instead of punishing teams for losing a guy to injury or just having a bad game forcing in multiple relievers.
Not really. Just make exceptions for both of those cases. If extra innings, you can use as many players as you want, and if an injury occurs, you can always sub for the player.
I hate this idea. This basically just means the dodgers would have 15 more pitchers on the payroll, rotating which 11-13 of them were active on any given day so your relievers are always available, unlike now where if a mop-up guy pitched 4 innings yesterday he’s out for today, or if you have a 19 inning game and your bullpen is decimated.
Add in the 10-day DL games and the fact that you only need 1 active starter and now you basically have 13 + 15 + 4 + DL available pitchers which would make games crawl.
Sure the 40-man roster would prevent teams from going that extreme, but the end game of this is always going to be more pitchers and more specialization which is the #1 thing slowing games down.
As is mentioned below, the 40 number might not get too crazy as long as you still have to use this same 40 man roster to protect promising players from the rule 5 draft or wavers. Say you want to protect 5 position players and 3 starting pitchers in AAA, while still giving them regular playing time. So, that’s 32 players at your disposal, making an additional 7 over the 25 man, plus those 4 and the DL. I agree that’s a lot. That’s a whole new bullpen every game. Now maybe you’d want to protect more players, since that marginal reliever really isn’t that good, while the 9th best guy in the minors getting regular playing time AAA is worth a lot more, but its probably somewhere in this neighborhood.
Maybe we should just let teams inactivate say 3 pitchers. This is almost always going to just be the 3 starters to have pitched most recently.
My issue with these proposals is it seems like we’re bending over backwards to try and get teams more pitchers, when we should be doing the opposite.
Why wouldn’t it be 30 new openings? 4 holdouts?
You are correct, it would be 30.
The other advantage of expansion is that young players do get valuable experience in travelling with the team, playing in front of large crowds, etc…..that has value to the players and the teams in coming years. Players get, what, lifetime health insurance if they are rostered even 1 game? I’d hate to see a few dozen people lose out on that every year….
while I generally agree I don’t like it, most choices seem to, once again, favor owners more than players.
Not unlike Proust in À la recherche du temps perdu, Sawchik occasionally experiences flashbacks to his youth — in this case, to that period before Colorado and Miami (née Florida) entered the league.
Also: duly edited. Thanks.
It feels like the Twins have gone all year with 14 relievers.
And they needed every one of them.
Do you think a 26th roster spot would affect how teams approach the Rule 5 draft?
Yes, that would update be the biggest change.
Those of you proposing the 40-man roster with healthy scratches do need to remember that, unlike the NFL, baseball has a minor league system that players every day, sometime in Las Vegas. It may be a good thing to give a lot of fringe players a roster spot, but teams may be in a position where they end up scrambling for veterans so that they can give prospects regular time.
This is exactly how it works in the NHL
Most guys that are healthy scratches in the NHL are either a) veterans that aren’t quite good enough to hold a top 3 line spot, b) younger guys that really have little to gain by playing minor leagues, and are at least picking up the game training at the highest level. You wouldn’t have top prospects languishing on the bench- they’d stay in AAA until they could contribute. This would really work best for young pitchers on innings counts and aging position players that would be considered AAAA. They could remain with the club all season and get activated for spot duty whenever. A lot of decent vets don’t want to hang out in AAA all season, so they have to make a decision as to whether to hang it up or ride buses.
We should keep the rules the same in order to allow players to make MLB debuts before the start of there first full season. Besides, most of the extra players don’t play unless your team is bad or mailing it in. In the case of bad teams these extra players are probably replacement level or worse. And in the case of good teams substituting league average with replacement level would actually make them worse, not better. So in totality expanded rosters make a marginal difference in actual outcomes.
I open the floor…
Would you really want to see, for example, half of NFL teams playing CFL-caliber players in the final month of the season, directly and unevenly impacting the playoff chances of teams who are still trying? I would not, and it’s the same for baseball. It damages the competitive integrity of the sport to encourage teams to treat games as exhibitions.
Actually, NFL teams that are out of the race do start playing their benches and bringing players up from their practice squad.
If anything, tanking is more of an issue in the NFL than it is in the MLB.
This still strikes me as a huge non-problem. The bigger mystery for me is why something I remember drawing pretty much near-zero attention and ire from the ’70s through the ’00s, has become Public Enemy #1 in the 2010s once the trade deadlines have left the news cycle. Even if I thought it was a problem, which I don’t, it seems like the 37th-biggest problem with the sport.
Pace of play and length of games, clearly. Time wasn’t an issue in decades past.
I think there was an article on this same topic not too long ago, and a lot of people agreed that it was a solution in search of a problem.
You may like teams playing random scrubs for a month, and directly impacting the playoff fortunes of teams who are still trying to win. A lot of people do not, however.
Having a 26th man would give room for a team with good pitching to have a hitter who was doing very well in AAA to get a chance to be on the roster. At the same time, I also like the idea of a 40 man roster with 25 active players.
Another writer suggested allowing the 40 man roster in September, but allowing only 25 players in any one game. That is meant to address the problem that teams that are out of the running, and don’t bother to spend the money to expand the roster, are at a disadvantage playing playoff contenders, all of whom are likely to expand their rosters.
Edit: I now see that other posters have suggested this.
I must be the only person who doesn’t have a problem with the current system. If you have more relievers, odds are they won’t be as good as your top guys. So there’s a lot more pitching changes, but from a competitive standpoint you’re likely facing inferior RP which leads to more runs scored.
Someone had a good idea in another post of having the 40 man roster but only 25 active roster (the rest being in a press box, like hockey), but you end up with the same situation we’re already in. The guys who pitched yesterday won’t be on the roster, you’ll have 7 fresh arms, and you can keep most of your SP off the roster to get more RP. So even if you have to set a “daily roster” of 25, you can still end up with 12-13 RP.
The problem isn’t that all 40 players can be brought up right now. Even if you limit it to 25, teams will demote SP or tired RP to ensure they have 10+ relievers.
And at the end of the day, I do like seeing prospects get an opportunity to play in September, something that might not happen as easily if the roster size was limited.
I think the system works fine, I just think there isn’t any way of getting around what most of you would consider a RP issue, because any fix or change to the roster just means teams will swap out guys to ensure they can do all the RP changes.
Designate a 25 man roster for each series in September.
This is the idea I came here to post and it’s similar to Moltar’s idea above, with the difference obviously being that you need to use the same 25 man roster for 3/4 days. You can “scratch” 2 or 3 of your starting pitchers from each series, and then change them out for bullpen help. It solves the issue of having a ridiculous amount of relievers available for a game, and also gives playing time to anyone the manager wants just by shuffling things around every few days.
If you scratch 2 SP you still end up with a 9-10 man bullpen. So while I like the idea, it doesn’t really solve any problems. There’s always going to be a ridiculous amount of RP available.
The biggest issue I see with designating a 25 man roster before each game (as a lot of people have suggested) is how do you handle service time? If a player isn’t active for the game that day does he still accrue a day of service?
I would assume they would agree that anyone on the “MLB reserve roster” or whatever you call players who aren’t active for a game as counting the same as now, where they get paid their MLB salary and accrue MLB service time.
Service time isn’t even the biggest benefit for September call-ups. They get a month of big league checks on top of it. That’s real money for a minor leaguer.
Right but it wouldn’t just be an issue for minor leaguers. Its highly likely 4 regular starters are going to be down everyday. That means if they don’t get service time while not on the 25 a starting pitcher would likely accumulate about only 6 days in September as compared to 30. That is a fairly substantial difference.
It has to be remembered that the 25-man roster is an artifact of a different time. As I understand it the rosters were set at 25 in 1908 (or 1914, depending on the interpretation). At the time teams essentially had as many starting pitchers as they had pitchers and could absorb pretty much as many innings as could conceivably be expected unless they were playing 6 doubleheaders in a row or there was a sniper in the grandstand. It was unlikely that even a truly terrible pitching staff would need more than 9 pitchers total. Every team had the luxury of mop-up guy would could eat innings when games got out of hand to protect the guys who were actually good or, at least, better.
We are now at a point, probably well beyond diminishing returns, of one inning (or less) relievers. In 1914, if you needed relief pitching you probably had at least 5 available guys, any of which could complete the game barring injury. Now we have 7-8 available guys plus several interchangeable pieces in the minors which are insufficient if the bullpen gets overworked the prior days and the manager refuses to compromise on the situational pitchers. If this was the case in 1914, a time when getting supplemental pitchers overnight would have been difficult, he roster limit probably would be 30 or higher. Well, assuming the owners would pay for it, which they almost certainly would not.
So to the point at hand, do I care that the Angels are carrying 14 relievers? That depends. How many of them are essentially a composite “mop-up” guy? Those I don’t care about. If they brought up a half dozen situational lefties, then that is super annoying.
I suppose the “real” solution would be to limit teams to X pitchers on the roster altogether with a designated mop-up position player or two who could only be used if the lead was more than 6 runs in the late innings. Teams would have to carry multiple long relievers to cover the middle innings. Less pitching changes, more logical bullpen usage, more bench players which I find infinitely more interesting than one-out-guys: what’s not to like?
What makes one think the Angels wouldn’t have 14 relievers if there was a 26 player roster?
I have mixed feelings about yet another reliever on the roster. Games run on too long, in part because the Manager spends 30 minutes on the mound. I like the idea of game day roster/healthy scratches, with maybe a September innovation of permitting the healthy scratches to be used if the game goes to extra innings.
Maybe one solution to games running so long is to reduce the time it takes to bring in a reliever. Does the manager really have to come out to the mound before bringing in a reliever? Does the new pitcher need more warm up pitches after already warming up in the bullpen? There shouldn’t be more than one or two extra minutes between batters when bringing in a new pitcher, barring injury. And to prevent funny business, any pitcher that comes out of game suddenly due to “injury” so a reliever can have lots of warm up pitches should have to miss a few games if he isn’t put on the DL.
26 man rosters sounds like an invitation to lengthen games.
Same with healthy scratches since that basically means all of your relievers are fresh every game, with the ones who pitched 2 days in a row being a scratch instead of just “unavailable”.
How about something radical to shorten games: 20-man rosters. Obviously the PA would never go for that. Maybe if they got a huge bump in minimum salary to offset the loss of jobs, or expansion to add jobs, or both, but really I know this is just a “what if”
Think about the strategy a team would have to employ with 20-man rosters. The multi-inning shutdown relievers suddenly become tremendously more valuable and you wouldn’t have the stupid Dellin Betances arbitration situations. Pitching changes would be substantially reduced. Much higher likelihood of a successful 2-way player since someone like MadBum might actually be more valuable in that role. Maybe teams experiment with starters on shorter rest pitching fewer innings per game. Maybe everyone tries to learn how to throw a knuckleball, or is forced to back off on the velocity to avoid injury (and thus increasing offense as well).
Or maybe a team chooses to keep all their pitchers and forego the defensive replacements and platoon PHers so they can shut down the other team’s bats but in exchange sandy leon has to hit in a high leverage situation.
It certainly would do away with the 9-pitchers-per-inning that everyone dreads right now.
You could do 20 active players for the game, and simply have teams carry a reserve roster for anyone who they don’t need for the game. Basically you have your starting 9 (10 with DH), and then have 10 or 11 bench/reliever spots. Which usually would be probably 4 batters and 6 relievers. Essentially, teams would be down a reliever or two from what they have now, but those would likely be the relievers who aren’t available for a game anyways, guys who pitched the last couple days.
or just allowed 40 guys in a clubhouse, but on 25 can actually come into a game
I am having trouble thinking of a reason why the MLBPA would have backed away from agreeing to expand active rosters to 26. There is no negative from the player side.
Because they probably doubt the owners will put new money toward the role, instead just dividing the money devoted to the bullpen and bench guys across 30 more players. Active players don’t want to share their salary with 30 more guys. That’s my guess.
The game really is not played much differently in September at all. Shitty teams would be mailing it in and playing minor league guys in meaningless games even if they only had 25 available players. Teams that are in contention have almost no difference in playing time in September.
Not true. Bullpen fatigue becomes a total non-issue, bc you let all your crappy minor leaguers eat all the low-leverage innings. Bad teams would not necessarily want to commit to being historically bad by sending all their bench vets to the minors. The current system not only allows but encourages use of both a high quantity and low quality of players. All so that a few prospects can log 50 ABs in the majors?
The 25 man roster limitation no longer reflects how the game is currently played. The solution is to treat pitchers and position players as two separate rosters- up to 13 slots for pitchers and 14 for position players- and require a team to designate 12 active position players and 4 pitchers (starter and 3 relievers) for a 9 inning game. One back up catcher may be designated before the game and need not take up an active slot on the roster before the start of a game. In the event of an injury concern, another 2 pitchers or 1 position player(all designated beforehand as emergency “first into the game” players) may be substituted onto the active roster during a game; with a removed pitcher required to serve a 10-Day DL stint and a removed position player being ineligible for activation (if not placed on the DL) for an agreed to X amount of days (say 5). In the event of extra innings, the whole roster would then becomes available to enter the game for both teams. Managers can declare which players are active when they hand in their lineup cards before each game. The same or similiar 40-man roster regulations would continue to apply across a season.
I think simply allowing a 28 man roster with three scratches(the other starters) would be best. My guess is that you would see more specialization on the hitting side. I would expect a three catcher roster enabling a lot more starters to be pulled in blow outs, Pinch hitting for catchers and the keeping of more offense-first backup catchers on the roster.
I also think you would see every team put a base running specialist on the roster and use him liberally. And likely a more defensive specialist for SS or outfield as late inning replacements. Possibly more three true outcome guys for pinchhitting as well.
I think just limit the roster for each series to 25 men. You can exchange various call ups each series and everyone gets MLB and service time regardless of if they dress or not. Let them sit dressed in the dugout, but if they arent on the pre-series roster list they cant play. Most teams will just sit the starting pitchers who wont pitch that series and replace them with call ups on the game roster
I don’t care what changes you make, I just like thinking of ways to exploit it. For instance, the 40-man roster/25 active plan. That would be a perfect opportunity for me to use one pitcher per inning every game. Pitchers are more effective in short bursts, but using them that way is impractical with the need for eating innings in a 25-man system.
“The MLBPA hasn’t been winning many meaningful battles lately.”
I would argue that the MLBPA was laughing all the way to the bank when they gutted the QO system that was in place. The new one that is going to be “felt” this coming winter is a joke, and is going to do nothing tangible to help teams that have all of their best players poached as they reach FA. This is just one very small step removed from completely unrestricted FA, which is what the MLBPA is trying to move the game towards.
I don’t love that there’s one month with a 40-man roster, but more importantly I’ve always been irked that it’s the FINAL month. Not because the games “mean more,” but because by that point in the season teams do have different goals: some who’ve all but clinched resting players; some gunning for a playoff spot and using tons of relievers and specialized runners, etc; some looking at prospects for next season; some simply gutted and done.
So it always seemed to me that the best and simplest solution is to make the 40-man roster month be April. At that time of year, every team is equally trying to win, so they’ll all use similar strategies. Also, at that time of year, teams tend to be more cautious with starting pitchers, so a sixth starter or additional relievers help more. And you’d get prospects up early with the chance to make a splash in meaningful games (not just spring training) and then stick for the year, which would be a much more fun story than seeing the kids come up at the end of the year when a team is already eliminated.
I’m fine with your idea. But if baseball is unwilling, how about you only get 8 non-injury subs (hitters or pitchers) a game? That, or a special Sept rule that forces all pitchers who enter an inning to finish that inning (injury excepted, of course)?
How about zero in-season rule changes, like every other sport in the country?
In the NFL, teams are trying to win up until the final whistle of a week 17 blowout, whereas in MLB, everyone is dialing it down for four weeks while the fans are supposed to be getting exciting for the playoffs. It makes no sense, and goes a ways to explain why so many teams look like crap as soon as the playoffs start. It’s essentially opening day after a month of “Fall Training” games.
NFL teams sit their starters all the time after they’ve clinched and if seeding is set. Doesn’t happen every year, but when it does there’s the same discussion of rustiness going into the playoffs.
I hate the idea of a 26-man roster. The toughest thing about managing a baseball team is handling usage of the pitching staff, not for one game but through a season. Teams have already punted this responsibility as much as possible by going with 13 pitchers. Managers who need a 14th pitcher in order to get through nine innings a night just aren’t savvy enough to be major league managers.
I like where this could be going. Most teams have a half dozen guys on their 40-man who are hurt, so you would be working with the following arithmetic: 40 – 6 DL – 4 of day starters = 30 players to construct a 25 man game day roster. 1 SP, 8 position starters, 1 closer, 1 stopper, 3 lefty specialists, 3 righty specialists, 1 long guy, 2 catchers (this is always an issue when pinch-hitting or swapping catcher, because, like, what happens if the backup gets hurt?), 2 defensive specialists, and 3 pinch hitters/DHs.
Specialization may be killing the game, so not sure this improves the problem, but counting those 4 off day starters against you seems stupid, especially if we are going to roast anybody who feels obligated to go short rest or extra innings circumstances.