Takeaways from Big Ten Softball's Sabermetrics and Advanced Statistics
Breaking down the conference's most valuable hitters and pitchers, including a surprising pitcher who rises to the top
CHECK OUT THE FULL SPREADSHEET OF STATISTICS HERE
In 2019, my friend Noah Coffman and I simultaneously had an idea: We should calculate Fielding-Independent Pitching (an Earned-Run Average predictor) for Big Ten softball. We could create advanced statistics, and then use them on our radio brodadcast of that year’s Big Ten softball tournament.
In our first calculation, Amber Fiser’s FIP was negative. Below zero. Something seemingly impossible, and yet nearly true. (I can’t remember what it was at the time after we adjusted the weights, though at the end of the season, it stood at 0.21).
That immediate and shocking finding led us to want to find more. It spiraled into more and more. By the time of the tournament, I had a full sheet of advanced stats, including an estimate for players’ Wins Above Average. We tried to explain the numbers on the broadcast. Hopefully our listeners understood.
After the season, I refined the numbers, did a bit more digging, and wrote up some of my big findings.
Now, in 2024, I’ve re-ran the numbers, fixed my spreadsheets and links, and adjusted the weights where necessary to give new estimates.
What’s in here is far from perfect. Something might have copied wrong, but hopefully I’ve avoided any errors. Some of the weights and estimators are, shall we say, a bit arbitrary. Then again, many of the official stats’ weights for MLB are just as arbitrary.
Other aspects, like park factors, or fly ball ratios, would be nice to have for more stats. They’re not available right now, but I hopefully can work towards more in the future. But, there are still insights to be gleaned.
A few more notes on this data:
All stats were taken from the Big Ten Website.
Stats are only for Big Ten games. Otherwise, the differences in strength of schedule would cause too much variation. I think it’s best to keep it as a closed system.
You’re more likely familiar with Wins Above Replacement. Here, average was used as the baseline. Average, in my opinion, works better for a college league, since “replacement level” doesn’t really make sense, as a team can’t sign a new player midseason, and what that looks like probably drastically varies from team to team.
For calculating Fielding-Independent Pitching (FIP), I could never quite come up with a right answer for how to change the calculation for 7-inning games. It’s also potentially less relevant in softball on the whole, though I would need more research to conclude that. Regardless, I sidelined it a bit, and used a modified version.
The calculations for hitters are based almost entirely on offensive data. Actual wins above replacement includes fielding and baserunning numbers, but without further data (yet), this will have to suffice.
I’m sure I’ll have more to write about this! There’s so much more in the data than just what I touch on here. And next season, I’ll be able to update it live throughout the season. Let me know what you’d like to see more of!
With all that said, here are some of my big takeaways:
CHECK OUT THE FULL SPREADSHEET OF STATISTICS HERE
Oakland’s World
In softball, naturally, pitchers often have a far bigger impact than hitters. They are on the field for way more of the season than any particular hitter, who gets just one-ninth of plate appearances. And compared to baseball, they can pitch a lot more than once every five days. That was less true this year, however.
Among hitters, one rose far above the rest: Minnesota star Jess Oakland.
Oakland alone was worth about 2.4 wins above an average player in her plate appearances. She buyoed Minnesota to the top total offense (157 runs scored) and a combined 3.7 wins above average.
Her wRC+ of 291, where 100 is average, means that she was about three times as productive as an average player. This also projects that the Golden Gophers scored 29 more runs than they would have with an average hitter taking her plate appearances.
Oakland’s traditional stats are just as impressive. In 23 Big Ten games, she had a .514 batting average, 11 home runs and 14 doubles. Simply, she dominated, and you don’t need advanced stats to see that for the Big Ten Player of the Year.
Michigan had the conference’s second-best offense, both by total runs scored (155) and by total WAA (3.4). Keke Tholl’s 11 home runs contribute a lot of value, while Ellie Sieler (173 wRC+) and Maddie Erickson (152) also had strong years.
Michigan also had the most balanced team, with 2.1 pitching WAA and 3.4 hitting WAA, both ranking second in the conference, leading them to the conference tournament championship.
No longer are pitchers as dominant, but the leader might surprise you
As mentioned earlier, Oakland was worth more wins to her team than all but four Big Ten pitcher. Overall, this year, the top pitchers were not worth that much more than the top hitters.
Back in 2019, Amber Fiser was worth 6.7 wins above average. Staggering. Danielle Williams was not far behind at 4.6 wins. Based on their total runs allowed, it was 6.0 for Fiser and 5.9 for Williams.
This year, no pitcher hit those numbers. Which is not that surprising. Fiser, in 2019, had 156 strikeouts, more than any pitcher this year, while not allowing a single home run in Big Ten play. Williams that year, as a freshman, had a ludicrous strikeout-to-walk ratio of over four. In 2021, Alex Storako struck out about half (!!!) of the batters she faced in a bonkers season.
Nobody did anything quite like that this year.
Ashley Miller, the transfer turned Big Ten Pitcher of the Year, was not the most valuable pitcher, by these calculations. That honor goes to Penn State’s Bridget Nemeth.
Nemeth was the Big Ten Freshman of the Year, led the conference in innings, and in games pitched (21, of Penn State’s 23 conference games). She gave up remarkably few walks, with a decent strikeout rate.
Based on the FIP numbers, Nemeth leads with 3.5 wins, and Maryland’s Courtney Wyche (the Big Ten’s leader in strikeouts) is second at 2.7. In the runs-based calculation, which will be used more in this exercise, Nemeth was worth an impressive 4.7 wins above average, likely single-handedly propelling Penn State into the tournament.
Northwestern was led by its pitching and its depth. Both Miller and Renae Cunningham had strong seasons in the circle. Cunningham gave up just two walks in 21 innings, and allowed a Big Ten-lowest .247 on-base percentage, ranking her as one of the most valuable pitchers despite limited usage. More on them in a bit.
Overall, though, the Big Ten had a slightly more offense-heavy environment, reflect by the less-dominant pitchers.
In 2019, the Big Ten averaged 4.79 runs per team per game. In 2021, that went down a bit, but was up to 4.85 this year. League-wide, the Big Ten had a .352 on-base percentage and a .449 slugging.
Was Northwestern Lucky … Again?
When comparing teams to their expected results, one that stands out positively is Northwestern. Again.
I wrote about a similar finding back in 2019 for the same team yet with no overlap in its players. And since then, they’ve pretty much ran the Big Ten, now winning the regular-season title in three straight years.
Northwestern went 19-3 in conference play, and went 4-1 in one-run games. Their total WAA indicates a team that should have won about 16 or 18 games. Their run differential also has them closer to 17 or 18. This is less stark than in the past, but continues a trend.
There’s likely a similar cause to the past. Northwestern’s defense seems to be quite good, based on what we can measure. Their batting average on balls in play allowed was .250, compared to a league average of .300. Some of that might be luck, some could be good pitching forcing weak contact. And there is likely good defense in that number, just as there was in the past (in 2019, it was .228 for Northwestern, easily the Big Ten’s best that year).
Oddly, Minnesota, despite three all-defensive players, allowed a Big Ten-worst .367 BABIP.
Pitching-wise, Northwestern had a total Pitcher WAA of 5.1, well above any other team. They allowed just 2.6 runs per game.
Miller’s stats reflect this trend for Northwestern. She vastly outperformed her FIP of 2.55 with a 1.61 ERA, leading to a 12-1 record. Northwestern as a team allowed just 58 runs in 22 games. Miller’s BABIP allowed of 0.229 likely partially reflects her success in forcing weak contact. But it also likely includes strong defense from her teammates. And a bit of luck, as well.
Based on the total runs allowed, though, Northwestern is right where it should have been. Cunningham and Miller were dynamic, with Cami Henry also pitching well at times.
There’s an additional factor of Kate Drohan’s coaching likely helping Northwestern squeeze out some extra value and wins at the margins.
The other explanation is that Northwestern has begun practicing dark magic. Who’s to say?
Penn State was both fortunate and unfortunate, it seems
To start with: I, as much as anyone, love to see the Big Ten get represented on the national stage. Penn State certainly deserved a tournament bid, and likely got in from a really strong non-conference record.
But in Big Ten play, they may have gotten a bit lucky.
Penn State went 12-11 in the Big Ten before losing to Maryland in the conference tournament. In conference play, they scored 95 runs (9th-most) and allowed 111 (7th-best). Finishing with a winning record despite being outscored shows a bit of good fortune, as well as their struggles in depth in the circle.
But other metrics, including their overall play, show that maybe they should have done better.
Based on Nemeth’s underlying stats, the Nittany Lions maybe should have allowed fewer runs than they did. Beyond Nemeth, their pitchers really struggled. The other three pitched 30% of the team’s innings but allowed 66% of their runs, leading PSU to just the fifth-best team pitching. While Nemeth contributed 4.7 wins above average, their other pitchers were worth -2.4.
Still, Penn State’s total WAA comes out to a +1.9, despite below-average hitting (it’s 1.2 in the FIP-based WAA).
Minnesota, meanwhile, should have done a bit better, and maybe a better non-conference season could have gotten them in the tournament. Both their total contributions and run differential suggests about a win better than they finished. And their pitchers’ FIP points to much better defensive outcomes and poor luck on balls in play, as seen in their .367 BABIP allowed.
Back to Penn State. On the other hand, their team BABIP of .288 was third-lowest in the Big Ten, and below the conference average of .300. This could be slower runners or slightly-worse hitters. Or it could be just bad luck, of the type from which they could bounce back.
This team could be even better next year. Especially if Nemeth takes another leap forward. Though they probably need another pitcher to take the load off of her. And the offense needs to step up. Watch out for the Nittany Lions next year.
Regardless, Penn State won the games they needed to win, and were rewarded for it. Perhaps that is the name of the game. In any U.S. professional sport, the top teams make the playoffs based on their record, regardless of their underlying metrics.