The Impact of Team Win’s on Saves

Tanner Scott signed with the Dodgers earlier this week and you may be wondering, does that make him a top 10 closer going into the 2025 season? Certainly, it must change his pre-trade auction calculator (Steamer) ranking of 14th, right? Roster Resource lists Scott as the Dodger closer for now, but Michael Kopech, Blake Treinen, and maybe even Evan Phillips may cut in on his save counts. And what if the Dodgers score so many runs with their high-powered offense that Scott doesn’t get the opportunity to earn a save often enough? Does being on a really good team limit a team’s save accumulation? The table below places team win totals from 2014 to 2024 into buckets and compares average win totals with average save totals within those buckets:
Though the data populating the bar chart above encompasses 10 seasons, only one team was in the “100+” bucket. The 2022 Dodgers did not have Shohei Ohtani, they did not have Tyler Glasnow, and they did not have yet have Roki Sasaki. They did, however, have a 35 home run hitting Mookie Betts, and four 20+ home run hitters in Will Smith, Freddie Freeman, Max Muncy, and Trea Turner. Of course, Cody Bellinger missed the party with only 19 home runs. It wasn’t all hitting either, these 111-win Dodgers had four pitchers who won 12 or more games in Julio Urías (17), Tony Gonsolin (16), Tyler Anderson (15), and Clayton Kershaw (12).
Win Bucket | Count |
---|---|
40-49 | 3 |
50-59 | 11 |
60-69 | 46 |
70-79 | 74 |
80-89 | 83 |
90-99 | 61 |
100-109 | 21 |
110+ | 1 |
The bar chart above indicates that a team’s save accumulation does not decline as their win accumulation goes up, that is, if we exclude the outlier 2022 Dodgers from the analysis. But the bar chart above does not consider a team’s production, both in scoring runs and recording outs. In order for a team to win and offer their bullpen no save opportunity, they need to win by four runs or more with regularity. So, let’s look at this question differently. I’ll create similar buckets over the same span of 10 seasons reflecting run differential, the difference between a team’s total runs scored and the total runs they allowed:
The 2022 Dodgers are yet again, the only team in the 300+ bucket. The rest of the teams show increased save totals with higher run differentials. There’s not enough here to suggest very successful teams are less likely to record saves, you could uncontestedly argue the opposite if you removed the outlier. On top of all that win power from their starting pitching, the 2022 Dodgers had 17 different relievers record wins. The saves? They mostly went to closer Craig Kimbrel who accumulated 22 of them that season. But, there were the Daniel Hudson‘s (5), the Brusdar Graterol’s (4), and the David Price–Chris Martin–Evan Phillips‘ of the world (each with two saves). Did the 2022 Dodgers’ high win total impact the number of opportunities Craig Kimbrel had to run out and record a save? Or, was it the strength of the bullpen and the many options the manager had to close out the game? Let’s look at the average number of relievers with at least one save per win bucket during this same period:
Win Bucket | Min | Average | Max |
---|---|---|---|
40-59 | 2 | 6 | 10 |
50-59 | 4 | 7 | 8 |
60-69 | 2 | 6 | 10 |
70-79 | 2 | 6 | 12 |
80-89 | 2 | 6 | 13 |
90-99 | 1 | 7 | 14 |
100-109 | 4 | 7 | 14 |
110+ | 12 | 12 | 12 |
Once again, the Dodgers appear as outliers in the 90 to 99 run differential bucket with only reliever on the entire team recording a save that season. In 2016 Kenley Jansen recorded 47 saves for the 91-win Dodgers, leaving a goose-egg for all of his fellows in the pen. What can we take away from the table above? The average number of pitchers who record at least one save is typically around six or seven. Teams with high win totals showcase a hint of a pattern, increasing in the minimum and maximum number of relievers with save counts. This makes sense as good teams should have more options, or plenty of good relievers capable of earning a save for their team. Don’t tell that to Kenley Jansen.
If you’re screaming, “Your sample size is too small, pal!” from behind your computer screen, I (somehow) hear you. But this analysis is also susceptible to the trends of closer usage over time. The way teams have begun rolling out relievers in the past 1o years is clearly showcased by high averages in the relievers with at least one save column above. Furthermore, this analysis is tricky because we’re viewing all of this data in aggregate. What if we simply looked at the correlation between run differential and saves in the entire dataset?:
W | L | SV | R | RS | Run Diff | RelCount | |
---|---|---|---|---|---|---|---|
W | 1.00 | -1.00 | 0.67 | -0.76 | 0.66 | 0.95 | 0.19 |
L | -1.00 | 1.00 | -0.67 | 0.76 | -0.66 | -0.95 | -0.18 |
SV | 0.67 | -0.67 | 1.00 | -0.58 | 0.18 | 0.52 | 0.12 |
R | -0.76 | 0.76 | -0.58 | 1.00 | -0.14 | -0.79 | -0.02 |
RS | 0.66 | -0.66 | 0.18 | -0.14 | 1.00 | 0.72 | 0.27 |
Run Diff | 0.95 | -0.95 | 0.52 | -0.79 | 0.72 | 1.00 | 0.19 |
RelCount | 0.19 | -0.18 | 0.12 | -0.02 | 0.27 | 0.19 | 1.00 |
The correlation chart above works like an old-school multiplication chart. Place one finger on “SV” in the header row and the other finger on the “Run Diff” index row and slide them into the table until they meet at the .52 correlation. Then, as you search around the grid, you’ll notice that the strongest positive correlation with saves is wins.
In the end, common sense seems to win this analysis. Draft good relievers on good teams and you’ll probably be ok. My hunch is that even when teams have high run differentials, the number of times they’re winning by four or more runs is likey still small, though that statement may warrant an additional study. We may not have team win projections on FanGraphs yet, but you can bet the Dodgers will have a high number of wins projected and newly signed Tanner Scott, with good luck and good health, will be waiting in the bullpen to close out many of them.
For fantasy purposes, how a manager sees his reliever’s roles is often the deciding factor in determining whether an individual player gets saves.
Maybe rank teams/managers with a Gini coefficent for save distribution?