Going Deep: We’re Drafting Saves Wrong
It’s no secret reliever usage is undergoing a bit of a renaissance.
The modern game seems to be doing away with the “proven closer” — much to the delight of Keith Law and Brian Kenny — opting instead for a bullpen full of filthy arms ready to pick up a save or serve as an “opener.” Whether this is good or bad for the sport is a question for a different article, but the reality is that bullpen utilization seems to be changing — and any fantasy owner drafting like it isn’t could be left in the dust.
Before we can adjust, I think it’s best for us to understand the crux of the issue. Is saying more relievers are getting more saves simply conjecture, or is there data to bear out the claim? Much to the chagrin of my wife, I spent my entire President’s Day weekend figuring out just that.
The point of this article is to answer four questions:
- Are more relievers getting more saves?
- Is this happening across the league or on an organization-by-organization basis?
- If this is organizational, which relievers are most and least affected?
- How do I adapt my draft strategy to get the most value out of a reliever?
This is going to be fun.
Gathering the Data
In order to best answer these questions, I gathered a lot of data.
I went to FanGraphs and downloaded the data for every reliever who got a save between 2008 and 2018. Each CSV was put into its own tab in a Google sheet separated by year. I downloaded all of the saves data for all 30 organizations in that same time period and added the appropriate information to each tab. In order to make sure that all saves were properly accounted for, I broke down the saves relievers earned on two teams. For example, at first glance, the data details that Roberto Osuna recorded 21 saves in 2018. For the purposes of this data set, I made sure to break down those 21 saves into nine for the Blue Jays and 12 saves for the Astros, marking whether a reliever recorded a save on the team he was traded from or to.
Next, I tallied the number of organizations with relievers earning three, five, or 10 saves as well as what portion of an organization’s relievers got 40% to 90% of a teams saves by year.
After cleaning and parsing the data and creating a few “totals” tabs, I was left with this monster that I’d love to share with you all for your own personal use. It may be nothing more than candy, but it took me quite some time to create. I hope you enjoy it: Pitcher List Save Data Sheet. Feel free to select File > Make a copy to sort and use for research.
Now that you know how it was gathered, let’s take an actual look at it.
Reading the Data
The line chart above shows the number of relievers who got at least one save by year. Let’s introduce some more specific numbers :
No. of RPs w/ one save
Total No. of RPs w/ one save
Let me clarify what may be a little confusing about the above chart: In 2018, 176 relievers got at least one save for an organization while 165 total relievers got one save. That essentially means that some relievers are being double counted, and for good reason. For example, Zack Britton earned at least one save for the Orioles and the Yankees, and so did Jeurys Familia between the Mets and Athletics. If we’re trying to see if saves are being dispersed throughout teams, it would seem wrong to just include Britton’s seven saves in 2018 as he is going to be utilized differently in different organizations. In addition, keeping this structure will help illuminate how different organizations utilize their pens.
For what it’s worth, below is a line chart showing the exact same info as above but using the total number of relievers featured in the far right column as the Y-axis:
The results are virtually the same and prove our overall point: 2018 saw more relievers get at least one save than at any other point in the history of the sport (or at least since the save has been introduced).
A valid first thought may be that this is some aberration. Maybe a perfect storm of injuries and ineffectiveness led to more one-off save opportunities in this past season. So let’s remove all relievers who just recorded one save for an organization and take a look at the percentage of relievers who earned more than three, five or ten saves.
Ninety percent of relievers who had at least one save with an organization recorded at least three in 2018. Seventy-three percent earned at least five. Both of those are highs in the past 10 years. While the 40% of relievers who earned 10 or more saves is actually the second-highest since 2016, this proves an interesting trend that we’ll touch upon later: The renaissance didn’t start this past year — it’s been burgeoning for years. In theory, you could make the argument that because more relievers are getting saves, that the threshold of getting 10 saves for a team isn’t what it used to be, which is a great point. So let’s continue to break the data down even further.
This is a big chart, and there’s a lot to break down. What we’re looking at is the percentage of relievers who got at least one save on an organization, accounting for more than a certain percent of saves for that organization. For example, in 2018, Arodys Vizcaino had 16 saves for the Atlanta Braves, who had 40 saves overall, meaning he recorded 40% of his team’s saves. Now that that’s — hopefully — a bit clearer, let’s see what the data says.
The amount of relievers who recorded more than 90% of their teams’ saves has declined from 5.6% in 2008 to 1.7% in 2018 with the lowest being 0.6% in 2017. The number of relievers recording 80% has also declined by about 5% across 10 years, as has the number of relievers getting more than 40%. This all supports the claim that saves are being more evenly dispersed throughout the bullpen and answers our first question — Are more relievers getting more saves? — with a resounding YES.
OK, that’s out of the way, but there’s a lot more to talk about.
Not only have we now proven that saves are being more evenly distributed than ever, we’ve also shown that this is no one-year deal. This past season was the third consecutive year the league has spread out its saves, and while the record set of 167 different relievers recording a save may not be broken this year, I doubt the end number will be that much lower.
If the league is changing how relief pitchers are utilized, then shouldn’t fantasy owners? On the macro level, it proves that we shouldn’t be chasing saves, and if the one thing you take from this article is that you should wait a bit more on relievers, then I’m happy. However, I think, by answering Question No. 2, we can begin to isolate which relievers are most affected.
Question No. 2: Is This Happening Across the League, or Is It Organization-Specific?
To do so, let’s further isolate the data to identify any potential organizational trends:
|Team||Grand Total||Grand Total < 3||Grand Total < 5||Grand Total < 10||Over 40||Over 50||Over 60||Over 70||Over 80||Over 90|
This data set is really noisy, and there is a significant amount of room for error in terms of interpretation and minimal room for nuance. For example, by looking at this chart, you can’t really tell that in the past three years, the Angels haven’t had a single reliever record more than 50% of the team’s saves. Nor can you tell that Fernando Rodney is responsible for the lone 90%+ the Rays have. There are still some very fascinating takeaways:
- In 2016 we see only three organizations — the Dodgers (Kenley Jansen), the Tigers (Francisco Rodriguez), and the Mets (Familia) — with relievers who had more than 90% of their teams saves. This is the first time this number dips below five.
- In 2018, seven organizations didn’t have a reliever with more than 40% of their teams saves: the Astros, Blue Jays, Brewers, Giants, Marlins, Orioles and Phillies.
- Before 2012, all 30 organizations had a reliever get 40% or more of their team’s saves.
- In 2018, six organizations had a reliever with more than 80% of the teams saves: Athletics, Diamondbacks, Mariners, Pirates, Rockies, and Tigers. Of those six, only three had a reliever with more than 90% of the saves: the Mariners, Pirates and Red Sox
Some more interesting takeaways from the table:
- The Blue Jays and Athletics are the only two organizations to have never had a reliever record 90% or more of their team saves in the past 10 years.
- Despite Mariano Rivero being the most dominant closer in baseball, in his last six years as a closer, he only recorded 90% or more of the teams saves twice (if we round up 2013, it was actually three times).
- In the past 10 seasons, there have only been five relievers who have recorded 100% of their teams saves: Kenley Jansen (2016, Dodgers), Addison Reed (2013, White Sox), Jason Motte (2012, Cardinals), Brian Wilson (2008, Giants), Trevor Hoffman (2008, Padres), Francisco Cordero (2008, Reds).
As we’ve seen above, the reliever utilization paradigm shift seems to have begun in 2016. If we’re trying to determine organizational mentalities, then it may be better to sort the data from 2016-2018 to get a better sense of which organizations are buying in and which may be more reticent.
|Team||Total Saves||>3SV||>5SV||>10SV||>40 by %||>50 by %||>60 by %||>70 by %||>80 by %||>90 by %|
One of the biggest takeaways I find from this data set is that only seven organizations have had a reliever get more than 90% of their team saves in the past three years: the Red Sox, Tigers, Diamondbacks, Pirates, Dodgers, Mariners, Mets, and Angels. We can tell from the above data sets that a reliever getting more than 90% of a team’s saves is rare, so let’s lower the threshold to 80%. After all, before 2016, this was much more common.
Across the past three years, there have been 13 organizations that have not had a reliever get 80% or more of a team’s saves. To me, this wraps up Question No. 2 — Is this happening across the league or on an organization-by-organization basis? — with the answer: NO. This isn’t a case where a handful of organizations are doing this with the rest sitting out. This is the entire league moving in this direction.
Once again, there is a lot to unpack in this data (and a reiteration to use the link above for easier viewing and more customization). As before, there are plenty of arguments you can make against this data set and its lack of nuance: Of course the Mariners had Edwin Diaz get more than 90% of their team saves in 2018; he was one of if not the best reliever in the game. My counter to that is then why didn’t Blake Treinen get more than 90%? Why did Josh Hader get 24.5%? If only the most dominant relievers get all the saves, why did Shane Greene get the fourth-highest percentage of team saves in 2018? Why is Raisel Iglesias not even in the top 12?
Question No. 3: Which Relievers Are Most and Least Affected?
In order to start to answer Question No. 3, I’m going to put the name of the reliever slated to get a majority of the closes next to the organization, along with that relievers NFBC ADP among all players and other pitchers (Jan 1 – Feb 19; Draft Champions):
|Brewers||Knebel | Jeffress | Hader||137 | 305 | 105||51 | 124 | 39|
|Braves||Arodys Vizcaino | AJ Minter||189 | 254||73|
|Marlins||Steckenrider | Romo | Conley||228 | 555 | 523||91 | 231 | 213|
|Twins||Blake Parker | Trevor May||360 | 238||141 | 94|
|Royals||Boxberger | Peralta||511 | 301||208 | 119|
|Giants||Will Smith | Mark Melancon||788 | 424||70 | 174|
Let’s all take a step back and appreciate that even Roster Resource (the source for Column No. 2) is acknowledging that almost half these teams don’t have set “closers.” Also, if you look at this data and think, “Well now that the X organization has Y closer, they’ll DEFINITELY have a reliever in the top 80%,” I hear you, but I think you’re ignoring the 2,000 words and countless pieces of data that came before this.
Of the 21 pitchers listed above, 13 are going among the top 100 pitchers off the board. By taking Corey Knebel at pick 137, you’re overlooking productive players such as David Peralta (138) or Billy Hamilton (150), who won’t lose homers and steals to other players on their team. Instead, you could wait for Jeremy Jeffress nearly 200 picks later, who could theoretically give you more saves than Hader and Knebel. Knowing the above, why would you waste a top-80 pick on Aroldis Chapman? Especially when you consider that he’s never gotten more than 80% of the Yankees’ saves. Now Adam Ottavino, Zack Britton, and Dellin Betances are there, wouldn’t manager Aaron Boone want to play matchups or be worried about Chapman’s command? Are you sure that because Osuna is an Astro from the beginning of the year that for the first time in what will be the past four years, the Astros are going to give him more than 80% of the opportunities? In that case, Hector Rondon would like a word with you.
There are other relievers not listed above whoI think could be impacted by the way the league is heading:
- Treinen, pick 64: To reiterate, Treinen didn’t get more than 90% of the Athletics’ saves last year. Now Rodney, Joakim Soria, and Lou Trivino (who recorded 9.1% of Oakland’s saves himself this past year) are behind him and can all get saves. Does Treinen have the best stuff of those three? Far and away. Would I be surprised if he lost five to seven saves to those guys over the course of the year? Not at all.
- Brad Hand, pick 84: Cody Allen was the set closer in Cleveland for 4.5 years. In that time, he never recorded more than 90% of the Indians’ saves. Granted, the Indians bullpen isn’t loaded like other teams’ are, but if I still think Hand can lose some saves.
- Iglesias. Pair the fact that he only recorded 63% of the Reds’ saves opportunities with the fact that the organization plans on using him when necessary and you get someone who shouldn’t be going in the top 100 like he is now.
On the other hand, there seemingly are organizations that are either willing to continue ascribing to the “proven closer” mentality or just don’t want to hear it from their fanbase when the guy they traded for isn’t getting opportunities. Guys such as:
- Diaz. While Familia, Robert Gsellman, Seth Lugo, and Justin Wilson all have saves under their belt, I don’t see the Mets sending anyone out there but Diaz unless it’s because he’s pitched on consecutive days or has become ineffective.
- Felipe Vazquez. The Pirates have proved reticent to adapt to the new way of thinking. With that said, Richard Rodriguez and his dope fastball are waiting in the wings, as is Keona Kela.
Personally, I don’t know if I’m a believer in either of those guys when it comes to repeating their 2018 seasons, but for those who do believe, I’d be remiss if I didn’t say that I think their value shouldn’t be as affected by the above.
We now come to Question No. 4: How do I adapt my draft strategy? Before that, let me clear the air a bit about the things I am not saying in this article. I’m not saying Hader should go undrafted. If your strategy is to load up on relievers to keep your ratios down, then the above doesn’t disprove that as a valid strategy. I’m not saying don’t draft these guys; if they fall to the right value, then you should by all means take them. My point is essentially the answer to Question No. 4: Saves are likely going to be in abundance more so than ever before, and you should adapt in one of two ways:
- Wait on saves.
- Change to a saves plus holds league.
To simply say “X player is a great reliever on a fantastic offense that’ll give him a lot save opportunities” is the old way of thinking. It’s taking the four questions we just answered, the piles of data, and ignoring them.
It’s 2019. Major League Baseball is adapting. It’s time we do the same.