How Much is Jim Johnson Really Worth in Fantasy
Jim Johnson has 42 saves. You own closers for saves. Ipso facto bingo blatto, Jim Johnson is an elite fantasy closer.
— or —
Jim Johnson has 36 strikeouts in 58.1 innings. The average closer would have 65 strikeouts in that amount of innings. Ipso facto dingo datto, Jim Johnson is a poor fantasy closer.
Seems clear that your opinion of Jim Johnson is going to depend on your beliefs about the role of a closer on a fantasy squad. If you’re just chasing the saves, he’s been a top-three closer this season. Some fantasy rankings, based on valuations, have him as the second-best in the American League. If you’re trying to look forward, or base your beliefs on the value a pitcher brings to all five categories, then your rankings might look different.
So it made sense to ask Zach Sanders to run his valuation method on the closers to this point, just to get a real sense of how much value JJ has brought to fantasy squads. Sanders ran his calculator — which gives players credit for how many standard deviations above and below the mean they are in each statistical category — on all relievers with more than five saves. Already we’re making a decision here, since we could have run it for all relievers, and instead we are limiting this to relievers with saves. You might be surprised where Vinnie Pestano shows up on the list he compiled with 1+ save pitchers (hint: 21st).
In any case, here’s your backwards-looking valuation of all the closers with 5+ saves so far this season (5×5 Roto):
| 1 | Aroldis Chapman |
|---|---|
| 2 | Fernando Rodney |
| 3 | Kenley Jansen |
| 4 | Craig Kimbrel |
| 5 | Jason Motte |
| 6 | Ernesto Frieri |
| 7 | Joel Hanrahan |
| 8 | Ryan Cook |
| 9 | Tom Wilhelmsen |
| 10 | Jonathan Papelbon |
| 11 | Rafael Soriano |
| 12 | Huston Street |
| 13 | Tyler Clippard |
| 14 | Joe Nathan |
| 15 | Santiago Casilla |
| 16 | Sergio Romo |
| 17 | Greg Holland |
| 18 | Jim Johnson |
| 19 | Casey Janssen |
| 20 | Grant Balfour |
| 21 | Steve Cishek |
| 22 | Sean Marshall |
| 23 | Rafael Betancourt |
| 24 | J.J. Putz |
| 25 | Chris Perez |
| 26 | John Axford |
| 27 | Glen Perkins |
| 28 | Jose Valverde |
| 29 | Jonathan Broxton |
| 30 | Addison Reed |
| 31 | Jared Burton |
| 32 | Alfredo Aceves |
| 33 | Brett Myers |
| 34 | Javier Lopez |
| 35 | Carlos Marmol |
| 36 | Dale Thayer |
| 37 | Javy Guerra |
| 38 | Matt Capps |
| 39 | Scott Downs |
| 40 | Frank Francisco |
| 41 | Brandon League |
| 42 | Mariano Rivera |
| 43 | Heath Bell |
| 44 | Henry Rodriguez |
| 45 | Brian Fuentes |
Yup, so there’s Jim Johnson, hanging out around the half-way mark. For every save he gets you, he’s costing you a strikeout, more or less. When you cede five strikeouts a month to the average closer in a category, that hurts. Since most leagues require you to start relievers, and since most leagues require you to go after saves in some fashion, then it makes sense to compare Jim Johnson against other relievers. And if you had three Jim Johnson types and I had three Joe Nathan types, I’d have almost 100 more strikeouts than you.
Relievers pitch fewer innings, and their role is less entrenched. At first, this seems like an argument for spending less on them in the draft, not one that helps us suss out Jim Johnson’s value now. But an iffy closer can lose most of their value with one decision by the manager. You do want to give Jim Johnson a little more value than, say, a Sergio Romo, whose ratios and rates have pushed his lower save total into a similar ranking. Romo is more likely to lose his job tomorrow than Jim Johnson. But when you’re looking back at pure statistical inputs, Jim Johnson has been about as valuable.
Since it’s more likely that a pitcher will continue to put up his career strikeout rate than it is to depend on a pitcher for saves, it makes sense to gravitate towards better pitchers — ones with better strikeout and walk rates. Even if a good strikeout rate is not actually a predictor of future saves, at the very least you got strikeout value while your pitcher was in there.
And so it becomes obvious that the future becomes a part of any valuation of a closer. Can he keep his job? Will he continue to get the same amount of save opportunities? These are as relevant to an owner as the correct evaluation of his contributions to date.
But we’ve seen here that it’s very hard to predict future save opportunities. The only correlations we could find were with team offense and team bullpen strength, and in both of these cases, the Orioles come up average or worse. Maybe you can give the Orioles’ bullpen a little bit more credit for their work in recent weeks, but their offense is not above-average.
We must give Jim Johnson credit for job security. There’s no way he’s losing the job now. Then his second-best asset is save opportunities, and his rate going forward is unclear. Since the average closer (top 30 by saves) has a 2.91 ERA, and Jim Johnson has a 2.88 ERA, he doesn’t add much value there. And then there’s his strikeout rate — worst among closers — that saps his strength quietly.
Add it all up, and you’ve got a third-tier, above-average closer — a ‘safe’ pitcher that should keep his job but needs a strikeout rate caddy.
[Update. I had Zach re-run his pitcher universe valuations. This includes the starters, and compares relievers among themselves and among the pitcher universe. It doesn’t value the strikeout and the save as equals from a reliever, but that isn’t to say strikeouts from a reliever aren’t useful. As you can see, JJ rises to eighth, which is about the highest I’d put him, especially considering he derives all his value from the save stat, which isn’t very predictable going forward. I’d still make him a third-tier reliever, but bottom of the second tier wouldn’t get much of an argument from me, especially given this list:
| Num | Reliever |
|---|---|
| 1 | Aroldis Chapman |
| 2 | Fernando Rodney |
| 3 | Craig Kimbrel |
| 4 | Jason Motte |
| 5 | Jonathan Papelbon |
| 6 | Kenley Jansen |
| 7 | Rafael Soriano |
| 8 | Jim Johnson |
| 9 | Joel Hanrahan |
| 10 | Joe Nathan |
| 11 | Ernesto Frieri |
| 12 | Tyler Clippard |
| 13 | Tom Wilhelmsen |
| 14 | Santiago Casilla |
| 15 | Casey Janssen |
| 16 | Ryan Cook |
| 17 | Chris Perez |
| 18 | J.J. Putz |
| 19 | Rafael Betancourt |
| 20 | Sergio Romo |
| 21 | Grant Balfour |
| 22 | Jose Valverde |
| 23 | Darren O’Day |
| 24 | Greg Holland |
| 25 | Glen Perkins |
| 26 | John Axford |
| 27 | Jonathan Broxton |
| 28 | Wilton Lopez |
]
With a phone full of pictures of pitchers' fingers, strange beers, and his two toddler sons, Eno Sarris can be found at the ballpark or a brewery most days. Read him here, writing about the A's or Giants at The Athletic, or about beer at October. Follow him on Twitter @enosarris if you can handle the sandwiches and inanity.
But if you count walk rates (which all good leagues do) for pitchers – I believe his ranking would go up quite a bit.
WHIP is accounted for in the rankings. It’s a 5×5 ranking.
How about in a league with uncapped innings (note: this is the default for many leagues)? The K/9 or absolute strikeouts are less important because a team can rotate SP through their lineup spots to get a ton of innings. What becomes more important is getting those K’s without killing your ratios (which is what streaming would do, theoretically). Then, since a guy like Johnson is a big positive in saves, and chips in a few K’s with pretty good rate stats, I would think he jumps up a bunch of spots.
I don’t think I’m in a single uncapped league, but I could see him being more valuable in those leagues. You’d still get better results if you maximize every inning, I’d think, but yeah.
But if you pair Johnson with any of the many many many excellent late-inning relief guys available (Marshall, Hernandez, Adams, Robertson, etc) you get all the saves and excellent k numbers from two players that were probably close to undrafted. Elite category contributors can almost always be accommodated with creative roster management in a way that helps your team.
Sure, not saying he doesn’t have value at all! But, even your response lines up with my saying he needs a caddy.
This!
Although as stat nerds, we love our ranks or tiers (although I think it’s a bit of a false dichotomy to say Johnson has to be lumped in “elite” or “below average” territory)…I think the above comment raises the more important roster construction question: How do you build a successful team/bullpen where you don’t “pay for saves”?
It’s inconceivable to me that anyone would pay for saves when you can build a top bullpen for $5 and a reserve round pick or two. I mean, there were industry expert leagues where Fernando Rodney went undrafted (http://www.mockdraftcentral.com/draft_grid.jsp?id=189028).
Further, look at the percent of 1st place teams on Yahoo that paid for saves–whether it be Kimbrel or Papelbon or Rivera (the big 3 coming into the year). It’s not high and it makes perfect sense when you factor in the opportunity cost of using a top 10 pick or $25 dollars on a closer.
All of this to say–if you take JJ on any RP for $15 bucks next year–even if they duplicate their success, you’re asking for pain…
And therein lies the problem.
If you compare relievers only with each other, you will massively overrate the value of strikeouts, because the SD for reliever Ks is far lower than the SD for overall pitcher Ks.
An accurate valuation has to take into account all pitchers, not just relievers.
How do Johnson and Nathan compare when you do that?
Or, to take a simpler way of looking at it:
Nathan has 31 extra Ks; Johnson has 11 extra SVs.
On average, how many points is 31 Ks worth in a 5×5 league?
On average, how many points is 11 SVs worth in a 5×5 league?
I haven’t played true roto in a while, but my guess is that 11 SV has far more value than 31 K, contrary to what your analysis concludes.
If you have RP slots and innings caps, it’s absolutely appropriate to run relievers as a group. You’re forced to pitch relievers, so I have to compare my relievers to your relievers.
And if you think about 31 strikeouts in a relief staff of three or four pitchers, that’s a big deal. Getting 31 strikeouts above average from a starter is not so easy.
Unless your league somehow counts “RP K” as a separate category, a K from a RP is worth exactly as much as a K from a SP by definition,.
So if your analysis tells you that a RP K is worth more than a SP K, then your analysis if flawed.
As, apparently, “if” my typing. *sigh*
Disagree. When you get 15 homers from a shortstop, it does, indeed, mean something different than when you get 15 homers from a 1B. Same thing here.
With all due respect, I am not seeing the logic.
I’m just using rough parameters, but let’s say that the average team in your league held 3 closers, with an average of 30 saves. Further, let’s say there was a 1,400 IP cap. with an .8 K/IP ratio.
In this case, the extra 11 saves would represent 12.2% of the average team saves total. The extra 31 K’s would represent 2.7% of the team’s K total. The save would have to be more important by a factor of 4x or 5x the extra Ks.
Or from a absolute numbers perspective, let’s say the best and worst finished 10% better or worse than average, If the average number of K’s was 1,120, and the last person was 10% worse, they’de have 1,008, and if everyone was spaced incrementally, then each slot differential would be ~ 22 K difference. In other words, the extra 31 K’s would be worth between 1 and 2 slots.
The same exercise for saves would have the average of 90 saves, with last being 81 saves and first being 99 saves, and the intervals in a 10-person league being 1.8, making the extra 11 saves worth about 6 slots.
Or put yet a 3rd way, if your league arbitrarily allowed you to add either 1 more save, or 3 more K’s to your total, which would you choose? One more save is far more likely to make a difference to the standings than 3 more Ks. Scale it up to 31/11, and it’ll be the same thing.
I’m not saying the save is worth the same as a strikeout! I’m just saying you have to compare a reliever’s strikeouts to those that you could get from relievers, not necessarily all pitchers. Considering the K/IP rates for relievers vs starters, it’s easier to get a competitive advantage with a closer with high K/IP than to do so with a starter, meaning that a reliever that costs you in K/IP has a negative value associated with him.
We’re re-running this with the pitcher universe, though, in order to get the full valuation. I wanted a shorter version just to look at JJ in the context of other closers. We’ll be doing all the positions at the end of the season, though, so might as well start early.
Eno, you said “Disagree. When you get 15 homers from a shortstop, it does, indeed, mean something different than when you get 15 homers from a 1B.”
I think you are confusing positional relative value with productive value. At the end of the year (or looking back at any point in time) 15 HRs are 15 HRs. Yes, I wouldn’t draft a 1B who i expect to only get 15 HR’s (i.e. looking at future value) because his positional expected value would be low. But when looking back in time, 15 = 15, no matter who gave you the 15 HR’s.
This same thinking applies to your question about Johnson, i.e are we valuing what he has done or trying to value what he will do? In valuing what he has done, I’d rather have the 11 more saves than the 31 more K’s. In the context of the leagues I generally compete in the 11 saves = 10% of the usuual year end total, while 31 K’s will only be about 2.5% of the year end total.
In valuing him for next year, I couldn’t bring myself to count on 40+ saves from him again, maybe low 30’s. Would I discount his value significantly because of low K’s? Not really, because I can find a SP to get me extra K’s, but I find it darn hard to find a SP who can give me saves.
I agree with Eno’s claim that you want to do rankings separately for RP and SP when you have separate slots for each type of pitcher, but I also agree that you want to use the combined SP and RP standard deviations to calculate the Z-scores. I do the same thing for position players, otherwise you get things like crazy SDs for catcher stolen bases.
A K contributes the same amount to your total regardless of its source, so you want to standardize the value based on all sources of K’s. But you want to set replacement value separately for RP and SP.
If you do not combine SP and RP for the SD calculation, then you also must think a SB from a catcher is about 5 times as valuable as a SB from an OF. And if you believe that, can I join your league?
There’s little use in saying that my shortstop got me 15 homers so it’s 15 homers, so I don’t think that “productive value” as you’re saying it exists in fantasy ball. Every piece of production is within a context if we are going to learn anything from it. I’m comfortable couching a player’s past production within his position because if we are going to put a ‘value’ or ‘dollar sign’ on it, that value number will have to be useful across all years of fantasy baseball, or at least useful to next season. Otherwise it’s a number with no context.
@Elias — yes, I did a short cut, but I have the rankings set up as you have decribed at the bottom of the article now.
@Eno-
Maybe another way to say it is that both pools make sense, and to try combining them? Relievers can only go in reliever spots (unless they’re Brett Myers/another SP convert, and have dual eligibility this season), but the cost is still coming out of your budget for every position. So while their production should be judged based on others at their position, the dollar value they produce should be judged based on what they contribute to their team stats as a whole. Essentially, you need to judge not only their production compared to who else could go in the roster spot, but what that extra dollar could have bought at other positions.
That said, I think it’s a case for looking at player’s value relative to the other relievers, and going off of what you need in your team/league.
I agree with that for sure, which is why I added the updated ranking.
Awesome. Thanks for the update, Eno!
Interesting to see Greg Holland, who has only been closing since August 1, at 17th on the list already, and ranked above Johnson.
Hmm. I’d say he’s been plenty valuable this year (saves are saves, y’all) but you wouldn’t project him to be nearly as valuable next season.
What was the point of this exercise? Jim Johnson went largely undrafted in all leagues I play in. No one who has him had planned to have him. Most people that do are the ones that drafted Madson before he went down, Thornton, or other guys who lost jobs as closers early in the year.
No one’s K/9 “plans” were ever hanging on Jim Johnson.
What he is providing is saves, and at great value. If your 2nd or 3rd closer only has a slight K/9 weakness and is not, say… giving up runs like K-Rod… you are already in great shape.
Correctly valuing a reliever’s strikeout contribution.
I just don’t understand targeting him as your example.
Jim Johnson was drafted as the 28th or 29th closer off the board in drafts? Even at #18 on your first list he is providing better value than anyone paid for him. No one who has ended up with Jim Johnson on their team at the beginning of the season was expecting his K/9 to suddenly skyrocket. It’s been poor his whole career. Next year, no one is going to expect him to start striking out more guys either and he will still go in the back half of closers in all drafts.
To me the more interesting name on this list is Valverde Where did his strikeouts go? Not only have his save opportunities suffered but his residual strikeout value is gone as well. No one here has written about that at all. That’s a man worth targeting a “hey you need to be getting a K per inning in fantasy” article on.
If you’re playing in a weekly league, I’d say he’s provided plenty of value in other categories not including saves.
His 2.88 ERA that you mentioned hasn’t been a consistent 2.88 ERA (actually 2.93 ERA). Anyway, he has 60 outings this year and his 58 best outings would give him an ERA of 1.25 with a .85 WHIP.
Basically, he had two terrible outings that make his #’s look a lot worse.
For those in weekly leagues, they’ve gotten elite value from him in every other category but K’s, excluding a couple of weeks in July.
With that said, I wouldn’t take him next year, which I assume is the point of this writing. Owners are going to value him for saves next year and there’s no telling he’ll have the same amount of opportunities next year (he likely won’t).
Thanks, it had been a while since I’d read somebody unnecessarily hating on the Orioles. What I haven’t read are any ‘Jim Johnson is the best closer in the AL’ reports on this site (or anywhere outside MASN). As the 20th closer off the board (probably lower in most of your leagues with his injury in Spring Training) he provided excellent value in relation to draft position. Who exactly are you refuting?
(sorry Tom B, I missed your post)
There’s a site that has him as the second best closer overall, and my rankings on RotoWorld elicited about the same feeling.
I don’t know why certain people hate that Eno used Jim Johnson as his case.
Is there another effective closer in baseball who doesn’t strike guys out? Jim Johnson is pretty rare. Just about every closer in the big leagues is striking out one per inning or close to one per inning.
It’s not every day you find a closer who’s been elite in real life with a below average K rate.
There really is no other example this year or in recent memory for that matter.
A value needs to be given to the fact that Johnson had an ADP of 210.8 (espn) and has given (now) 40+ saves. Any smart person that cared about strikeout rates would couple a closer such as johnson with a high strikeout non closer (see: Clippard before closing, David Robertson, Grilli, Hernandez, etc) Spending an early draft pick on Rivera is not a horrible bet, but using that pick on someone else and getting Jim Johnson more than 150 picks later is something that NEEDS to be looked at.
ALL THAT BEING SAID
This is well done in the scope of a 5×5 valuating the performance of a closer within the categories they effect.
I really thought I looked hard for this argument. Guess not. Disregard.